diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/checkpoint_metadata.json b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/checkpoint_metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..2083a6c0e3511a9971b26486110afd01b0ccd94e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/checkpoint_metadata.json @@ -0,0 +1,18 @@ +{ + "custom_metas": null, + "dp": 2, + "metas": { + "consumed_train_samples": 512000, + "data_stages": [ + { + "consumed_train_samples": 512000, + "name": "stable phase", + "start_training_step": 1 + } + ], + "last_stage_idx": 0, + "last_train_step": 1000 + }, + "tp": 1, + "version": "1.4" +} \ No newline at end of file diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/config.yaml b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/config.yaml new file mode 100644 index 0000000000000000000000000000000000000000..0d2fa5a03fe2809e896f81c13cbff0d6d34d0743 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/config.yaml @@ -0,0 +1,107 @@ +checkpoints: + checkpoint_interval: 1000 + checkpoints_path: checkpoints/llama-3.2-3B-finemath-4plus-finetune-bs512-60B-local + checkpoints_path_is_shared_file_system: false + load_lr_scheduler: false + load_optimizer: false + resume_checkpoint_path: checkpoints/nanotron_pretrained_checkpoints/Nanotron-Llama-3.2-3B + save_final_state: true + save_initial_state: false +data_stages: +- data: + dataset: + dataset_folder: + - finemath/tokenization/finemath-4-plus-p0.0 + dataset_weights: null + num_loading_workers: 0 + seed: 8 + name: stable phase + start_training_step: 1 +general: + benchmark_csv_path: null + consumed_train_samples: 512000 + ignore_sanity_checks: true + project: llama3-3B-CPT + run: finemath-4plus + seed: 6 + step: 1000 +lighteval: null +logging: + iteration_step_info_interval: 1 + log_level: info + log_level_replica: info +model: + ddp_bucket_cap_mb: 25 + dtype: bfloat16 + init_method: + std: 0.041666666666666664 + make_vocab_size_divisible_by: 1 + model_config: + bos_token_id: 128000 + eos_token_id: 128001 + hidden_act: silu + hidden_size: 3072 + initializer_range: 0.02 + intermediate_size: 8192 + is_llama_config: true + max_position_embeddings: 4096 + num_attention_heads: 24 + num_hidden_layers: 28 + num_key_value_heads: 8 + pad_token_id: null + pretraining_tp: 2 + rms_norm_eps: 1.0e-05 + rope_interleaved: false + rope_scaling: + factor: 32.0 + high_freq_factor: 4.0 + low_freq_factor: 1.0 + original_max_position_embeddings: 8192 + rope_type: llama3 + rope_theta: 500000.0 + tie_word_embeddings: true + use_cache: true + vocab_size: 128256 +optimizer: + accumulate_grad_in_fp32: true + clip_grad: 1.0 + learning_rate_scheduler: + learning_rate: 5.0e-05 + lr_decay_starting_step: 50000 + lr_decay_steps: 10000 + lr_decay_style: linear + lr_warmup_steps: 1000 + lr_warmup_style: linear + min_decay_lr: 0 + optimizer_factory: + adam_beta1: 0.9 + adam_beta2: 0.95 + adam_eps: 1.0e-08 + name: adamW + torch_adam_is_fused: true + weight_decay: 0.01 + zero_stage: 1 +parallelism: + dp: 2 + expert_parallel_size: 1 + pp: 2 + pp_engine: 1f1b + recompute_layer: true + tp: 1 + tp_linear_async_communication: true + tp_mode: REDUCE_SCATTER + tp_recompute_allgather: true +profiler: null +s3_upload: null +tokenizer: + tokenizer_max_length: null + tokenizer_name_or_path: meta-llama/Llama-3.2-3B + tokenizer_revision: null +tokens: + batch_accumulation_per_replica: 32 + limit_test_batches: 0 + limit_val_batches: 0 + micro_batch_size: 8 + sequence_length: 4096 + train_steps: 28612 + val_check_interval: -1 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/lr_scheduler/lr_scheduler_pp-0-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/lr_scheduler/lr_scheduler_pp-0-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..c6987d39f10a06c9899c9aed6d4297b1d401d524 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/lr_scheduler/lr_scheduler_pp-0-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64b1904c41c918754e15d16fc0e1250ba24c7d568b08c27ea9438c93bf43e42b +size 3248 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/lr_scheduler/lr_scheduler_pp-1-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/lr_scheduler/lr_scheduler_pp-1-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..a45c26135834dd7dd19f0033d7c300e89f9e8329 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/lr_scheduler/lr_scheduler_pp-1-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01127dadd93637eddd730e5822267d7a3ce61f07f6c03ac9873b3bd091ab23d1 +size 2800 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6c5afddd5efc9bc79ca4524a11bfd407befdeffd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5493674de13ce7e3df04d0d63ff63e66be798c68335816a59ea613ea0ccc1167 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d642ae0b6acb11e92f239208f7594f4178e1dff7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aa86799990558ccb70dbfc61e66689c50920545d9e3f3a6764adb6b32eaa7ecc +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2d226f8e9bff6f411a9979783b72ed617e235be3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59c192c095b876e2b7ca46a6ffd8de0095a18b334f79b5e5e50c5272a1de7186 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cee17d0e154b5039fdc5549e38d88c3133fdd85a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f77148dd47d7f06a7b11def2f0734dd56803a674da56e7d88fb52cde0adb688 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a85c4199bdd85d4580a673b9ebae03796a60c438 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de47a53a6b40e494bb634849a5dbd37154419e4c69dcc33adbb949722e956cac +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..39bb5231c45c00e58baf839b058134671065a5da --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/0/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:879e858211fac1a5dc714b4ce97e073b987ffdaa61dda9f19873971347ca1bf8 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aaccf6753f067aa117f03057326858ac915c4f2e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ead6cb8f4d40c5ed60d35fda7a4f5cd16abdf0c172b8f24623b11d7ba4b93817 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4314c6c56ce0658d6ccd8edc8cc678d8c2211d53 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52c07408036d849db0dd0ce59a77dc8ef9a904ff5eba85a1b81ce02ded24c87d +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..555784241ead09bef6c3b392f0148fb20b5451ee --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8254fea25c9490957bd8ea365412df679ce42f4b479fba76a0a40805d03f944f +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..94dcf8025269d1d293500862d67a578593ea6e6b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:30ed749a776132071616e7619cb1cbf5db19b050e7bb305778fbb5a6181d8c09 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..155b7f8165accb30cfff3e1fec627873c4c1dcb5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b2cf107f041963311af5f8dba437ed0d04b3857cab75e757bfb97bd9e2c6cac +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..23ca4df4a36d6609679ff0aaf5ab5543d29e193b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/1/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe35fcb4da15a277088739e457da9cf6967a5a3ffcb71bb26b55882ff4fafb7b +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d1e6bac51dec4c977b50f10db9dae5264f01acfc --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ff3bd80a078901359e3fe781f03873e4912568fcfb884a75724e7039d3e6ea76 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c32bef84fa1ffcf0710f60ac62324f365ac5092d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f3b852d2ca2cd436f56c589befab59f564b4c3d4ce54e45560f02fcfa86753bc +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..81e0aed8df8ad2e2ce582d32360aca93ce7936f1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9fb8a20d14a6333756b3d1767f911b18649c3bf6b7f00ff7457e651e0c581f5 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8e35e5ef61f1bf1cbe99a8f454945307307e00e8 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:970e4220d7e388b58243c575ed919de5192c44e7abb5981aaed3bc0ffe8a9c3e +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..52b764fdb136d9a1bdc61fdc4549cb13d90b6043 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35b6453d359d80a167deed46d2689df359566c730844065584fa3ac4769158d3 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9d88a150cc20e4c3fbea7034bd91c68710a4f969 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/10/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b76971f8857c2ed4efa0e4a52eaa05306e5e7e0690b280d644e74d80d18da79a +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1a860999c022f81e80f1a56fef329a4a30d68e38 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a349da7992ac30258073cdebdad91a4c4055f78406643837e24b01eb2e0124d +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e86dfda9dcae359e59be46339cba86d0a62a925c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7e61876d53a05d71b54d61895e3070d2dd158f467762cc07710f52eaad56230 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..56baa0f104f03b3e54002d4d11592bda8e404ba3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eab201b2906337f52643da1625302434ce236b2fbe5a07cc6edb987b6a599903 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5449cd4516402861e533551f90083d658e87010a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:568cdec1f52398f33c96bba87f6445b8c2133d5a55ba137050d5bc6d9f478b28 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..73894d81e33ce9a9935b07713743c54babda004b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c0617327ab4f4534272f67d43ef0bbb1f209565bba29187b5b94caccd75141d +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4493bd2c391b08b54c45fe29c685695b10327e91 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/11/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:58e20f32666687910414d6573c24eec2534fde07f1f233f40b2a10ac7109f749 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..61deaa3195887387cc240a8fe7e5be1f6d08fb03 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca530381500a4a04df3390580ac45fe365b98268114a3f1fb96f999cc0d7fad3 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..54dabd4c3cb1524bdd82a13944f06db8d44ca961 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cce4a333e0cf814f6aa0e4aa4c4ad9bce546acc48689ff42771211d417c9f2ce +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..46526d9c13b8bce18c21a63938dedb73f36e3487 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7489902f9b554936da00e2db9b2ee0e0e473c97f122b68cc8aa552773d2505a5 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0d05679d684d8915ae81617685bd12b21aefe1d5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22e8f42acb23876dd424d70d4c2c8f84532fdc14b51f672f5d9fb53d8c95ba0b +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..469cc55c291d77e1315f4fbe9cd6383cb0bc7d33 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc41a636d0fb672b49d5b50c596f82291a46957fb25e48492a4d29e2ed1ded1e +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d48eb1dc614079f34366429c261ce04adbb8dfaa --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/12/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f65928096d6874896b8e1a82f5c85d58b0b57f374927cee9aa8228162c57238 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0e503ef0f2e5dac49040ac6a32eda4dddb52eecb --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5663a2a3ffe435e88be4648ac7dbafd1b20a56068fcd1402674a7998f61215e9 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2fa895b668a8376e810798866ded6f3a309689ba --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:82c6cbc3afe11e568e4ec14c05d7b30d6d8cd23ef3027713e9f3d196d7c856f5 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5aa55d3b6dfbecf0215384f99d217a93d3a3b8c7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0a89b383e9b081be8456e4e19ee71f8c830fa41c0a9d7959555cbd37408f1e9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b8c23027103ac6e53678687c43905e9714a1bbda --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b19cec7cae68d67a0b21eb8b4d676cc567e3b633300b91c76af7d925f92d3bd +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ef725a1009c7ae7cdc723fd446af1f64ff074924 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d226d99e5f1a16c5e0ac6b751342de343678070b796f8aad6b98df5aba6e174e +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5ca0568fb31841964976505abfb3a619344a3b50 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/13/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d5700ed85925816302d8c8c1f8a0897ed40fab88a0543043a4bc9d0a0feb892 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2241a40518433a286130ac1ade2aa80df99b6c1d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a80aa90e167e1cab4b9c580e2acc28ce655857cb3ca8a76b7c4120ebaa607103 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2e9694b4d95a24f0cbd64ff3e0b037acc22d191c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c21335f74c5f5ecc31dfee4e9a9e2c4683428289d738db5dfa45c975e565923a +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d133037a6118a4d0f98799de6ca99bd7d48090c5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:222134760b958d6ca7022fc6f7a9f299311afa2e99b65502c379fc73db2e6596 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..345d98cfbe4fc7b24144da6ace26c50c1ee7d107 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d6a17fcc580b691a7efc41bc9bda98b5689d6b4d00e1df29f2767165a2796be9 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..859ab14f9a510c24001324cab09947da54d299d9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cd5db9eaaf927f14d1ef2f91b94a709194543e89d140812b2fed94faca896257 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6b9f7328ed10552fc72bc6c1e928b9066616dcac --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/14/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5decd84a8eff9d0394597070333b6faf03878dda3eab2145f81bb7426a6739f8 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c27e016ec43abe8472d5fad3acb27e2379c03b0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:54f03bee6326767a1d39dea40155e6f4c1dbfd84d25a603d54ee69fcaf91e40c +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..55e81013085f24b70e4a9a0173fde4f5c4a1e91a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d9cdec71c9db7e953f1fd7705024115346d92a7313028e78c3e62630503d28f +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..441663d9655bd84642dcce37cae725e01bb359eb --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d18fa9dec303593a03dd8404e793fbd5b5266e4b7941990e7d539d7a909968d +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..94ed79734739896fdb6a889ee310b9422ca2a11a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:456d75fb3c0bc293ccd8d6a508bc9df708ce5edc437e18cb5d9f207497ca0f85 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4eb4dfe231e23ef4a0ebf639da8de57085e77df8 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eefd17d76183b7fbf12409c2a43b3ace6d9b0b508d88f6a15792f79a2370ce9a +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c581a0e5e8008cc57d52e73eedfc0ca01a6a631c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/15/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f29746550ab77892cd20dfb0a23e655e32204971fb750f11f56ce316d4638c8 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..047bc69bdb9fda7f3abc2924d47e78ac123488d7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e5f8f7cde311d99a5e578e03e3f8fe58831ae89be715932379ca9a25c7cbc0e +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..afcc796bcf21ddf2b58514b7dc59db84013d75e9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d15282f61b7d5648cbfd4db9037a1ae871b737303c6c0fd19e01ea1faade03c9 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7d24ba9a1901c709e04f081fd8de642a6796c6e2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92e4d6ba97020bf54cf4b9465f9e28729279aaece07f5901fb106a597a46cfec +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c6a76239e7b9cdbdbd6c6ad9a4d1bc85e3b8ac28 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61e85964073d650498bdc14da8a42307890500d5b0156118a0e97c3ed9a8ced1 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..239a4f5c9e98b62323037a859bc1d18276b71008 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:454e483c24ec863c3505f11d6988716c4d770be2564068e9a78c456da3a20f60 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f44ae1f7439f7130ac3766103693a913bd019695 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/16/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49183e8d3643e5d999437b136fd2fa2b2d505acf3e5617ce607a13f5dad83e4b +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..415d523956f03a1b966f5d17f609d48d24677b59 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3bb4836b6199a1a1f535a029221f97e42ae5e62bcc861f43d8ddd58b61a55d0c +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0d30a2b582107c0268a9c5ffe8f5a9e92fb7631d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b5b29ab19575ed0b41453e9233253f55316f0684c4c6ea1cfe456fc7bdd7d25 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0c74cef67f20d482e4f94dd2950eaafcc0ebea77 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6aedffca3490cbd1d9ae3341103cb8a490093a653bb4e321ef19d5a78fe4c4b4 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b7982e814827671fa935b503a9eaa8afbfaa357b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3c21fa4010dec9af3ca47d9b47e14077a9f1ad0d295058073d50cf150650f2d +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ae06d84a80f4c2317c671694a3ad92bb3011e6b7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2f47a859fe902ce7dafcb7bfe939afbd7173aa54a8f430208d3debbd215a77f +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..96433059647db8a6cfd6cf6c8f530bcab4d91987 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/17/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe63e220ae00b9d4a9f313c5fac21143cfbe4580410f50789d5eed08520a2e98 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d3811acd5b12757b258e30ea8d5f2d7717b85411 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b61e5b44f934c22f9dc6bc19f4660859f4e940e8ca5fc0d679ee8455ed6c15c5 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..420c8341ba0abe114150a280a09bf1a56f642c41 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:81a87d359db03da4b4ebf7d75550bf04bf860088433828422b34a2182d47b514 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a93b0cd373ef6964409801014a71d6bc4c3e16c7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f7c530b022019205a7efaeaa27826acc617f4f4b569a872ad3e4879106b48e0 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b66b1093877077232dce9b352016b11b85c61a11 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:953bb44ab024975eb10f7e28296b3477c75073763957db8e889bd2b0614a6b51 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2b6035dfd49209b7a08d3dd410cdb01803d5e00c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84125fb50274eded83522acaf314b7df569900b8873ef82f679edd94978c93a4 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b67aa6f0c063453ac36b976805b6bea27cd9223 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/18/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcc35a00a6b2c0dcf7a5a63040760bf4266e2f4bb4723d2c99f53de21002595a +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b50f33da73519d241d35e9acc1ae00e1a06b0672 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b325bd077c47bf72ed086e7cc067226b440b711977116886ff6ea49cc432d42a +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e34f719fd3a6d792cc67b7bcc5321d6a77a0f040 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71c945d454ec3320305bc8f88fc01f4b76ee8a96357ffe4d24bea123434c11aa +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..12729726a6c609515cc4a631183a4889f3dffc73 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fab81fdce68e096139c2c7bf3f5a8245be650b4c675e017e40298f7a1add4ce9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cf23473d08a50ac1a45fada958c6584cbf8bc45c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5ba1197582c25f34926bd30f64ea6ce7828c6985344162a67863283033cd5843 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1ece92acb4dd7f4a3fa7f6c4257d32dc6bc59757 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e595aa8044df422a058f6ee3f84215979b7f431a3736b7e5ab9b745dc1815d4 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..57ac9388aafd509706e2bce66660e94d72f68470 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/19/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46d99af4dcfb73b312b807a16c00786b63c37fccb5d2f69913c41d174c370b03 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aa189b19a494b177023345d2bd189c50c025b43a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dba7e53550fc7fccc8ca6905084c8cfcf4791680ddd680884c4dfc75764b3726 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..35170d5034c420eb652544502b4b46a4e7cd0e7b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e7e9140f90205cc5580c345f9d1f516c0521132690a62e4b79510e161f1328b +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a6ac56025a733945190296a7cc21ed379c3b456b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89ca26472e36250008a06a5f4d10068e13fc706d03747e4bc8df66ffd23d5835 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b6dff733e0e899de8e077542a4de89a17ed07e24 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:039bbdb32a7921facaf307087c5420412176cdc30a311c7b11c274344fc7d766 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f76435f7042e0785278cdbc63507f7186bd71f2d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d91577555d4f7c4c3631879a3d34462c3848f779a208e50f5f9b7acad1a6dcf +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9018fcdd5b918e274a2cbe0759b89b7180a1bdc6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/2/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b7029ed6a82b7148900cd9c9a8ee68c03f7932fde45ebcc60b5678026dd40c9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c70bbe9b6465baa0603705268f7274787b2db4fe --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a1083c7d8b3dce0090b0d9398c08fa767ccedd67d9529b3541476e3065d10f3a +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..656dc9c8302269c7d735804ec33f4be5919ae922 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:171cf5b9e704e195d144eee7c26ef758c685a732f7fb7469054792d316e9cc63 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b613426dafb9a5502c560e096de3574fecf49da4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df68249cc966c77d38ffdeadf42366abce17113d031ca559872ab1ae48729d72 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ff2e4437da7c90496167341d9cc32b796591d5fb --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ef3d37fca6ea50e34e8fdd5924d4f53d6c41059dd6ff4b0652e7271ed44f346 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6e71b050d8d267b3b7a537d43331d54a11d61424 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b4666d06a1b28295f2f4fce76d0272453246e970e9d2ea6ebf112546579a174 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b0d7b6a5b5a3dd6d87b1ffdd323a9441b7b57d03 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/20/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70acbb7d5ab6c067d8932cbb426c38da20f48fa33a0c461c0c309d34b982b6de +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..891c00600219a103c1f7eb6635d4c38d3f008d09 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3c5b037312c032c9d6eb52879a2a327a0eddaecfd404f932abe41fa41df476d0 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f30fd1d3766836a5c2321a6f341665060b9f83d6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a835cc1ceac6c85ef5448aa154b911b7fc871134d23486db3cff4391548fa7b4 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..76bbb985c3be00fc215b499261da5c25bcd2989b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:257a98e517d5c309196372519933cc7605a3d3de41a3a9e568d03e882af2c1f3 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ae35a2f893941ded6a622062d2352f80d35c4aa --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2749323cb2bceae29ab68a7252ef45260685b2ca456bb74985dfcb40e18d898a +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ba275de24f9ac64fd2b9614822c50e27b0af2ee --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae103c06c6992cd1cdb9424ce254a95972ef5de737b195aba1dcab0ccebadf42 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..64a0cd53e7ee3157b1de90e00930537d695fefc0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/21/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1d566aaa14170db0ae365f2f481c84e32fb95edb86e2fe97311b3c8147818191 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..024f95f862c080a1bcd08defc7f742d2e68114f9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d0d25b80092e8ad6ce701606131a66c420051d73b4e6e339022cdc1365d6209 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..813018cdbfb76589cf772951ba0801cbd7e03edd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:07d0dc9def047d03b4bb2e470037e46787a9088576ba09513fe65bdb8222c2b2 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c104dd202d2d2ed66f06c15476393ab67742c4f2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aecd862d4d23c767de1880558d2f326f17d8b7a5ea54bacaab4ea3cbfcd583a2 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3245c10e69bf85320480c46c0378d2698be7fd7f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c501c9514e42d5b5029a65ba85608c374caad806a495c5dbcd81eb64f6ce00e +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..81ee31c224cdec063d5433b5b40d223d522dea2d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e28315d317a68dccd094d0ed97b5aec0d7ce8f6eecb886968d2b5a0f4bac1e3c +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..934e746c3b384c4fef1446d3c21fb4edbf8686e5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/22/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a028673dcd296193ffcd6e110c9b747bab2dfb856346a1754b3be7af0bca57e6 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cd4f0363904b4b5cdda02fe65c716b93ff12954a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:165e67530581f76881de41605ec3d77c7a843aef8a0c51be662dd6c1b524cbc6 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f8bd7469385daa62a401ac798a96b46d3ba35ca3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ae5bc8f7d86ff4f2b9ea1ae51fff855db3631e77aa05f8b19fa66c17564b3ae +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f1a6d5c9abb5da8b20a2ecac1e22d43cfa83bb50 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e7f399e6d150c0c12578044090269e15deb5edfe634174e75ce6b106c439a67e +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..dc8917fba8d4815446775c884d0d8d6b4bfb6c4a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b45220df6b1a8ba85ff97e481c674bda174567ac50c90b8b228ea034761a335b +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f5e8949a3f7253558d453506ca5d0b058782c593 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9acface4951877fb9355bd1b31f1438069211cad7237f155ae52b1e87a408a19 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6956a5b6d3b41470c3b3ffdf5b140195b389eae6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/23/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:214c5d5833cd2d38443b1ae35cfbaa35dd5fd97e88f867122148c24ebd0ca346 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..06980a8ae4398d282e0dd7d2bedc5b75bc9c4180 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:80a0dce0deee816c5ddaf90e029baf442564ec22bddbe2e6b9461f1bd88dfd3a +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..43423f96315240306e9a57aeead89b78891c5c7d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9ab8074ee4c4e9801ff751c55cb0f7a42fb4c27b0fc690c78c4ac2be92e3f12 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ee8bcb56cef3469d4699677dd40e485dd99012b4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:44dbcc9bdb9d1803a3571193875ed302d3809b59578221b72795929f5fec1ccc +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..05a5b7378a919f365dcbc145efa4134645b5afc9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:264a4188c0d65246eb5c8ec9200b387a860a94acd75c0391e8177b64916680d2 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..373f46a82c303f747e46e24d101157a07b9156f1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:140491340c335cd090d774cda1c5c49d261dc612f92aec2014cb27b08265f0f4 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bac2a3d6b4571123f2de0e73c453eaebee18dcd1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/24/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a58038d4abf8ca894ce8eda260f1839536f49c1f157030c5caacd1b0b1512127 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3b65d70cce3935dfb6e73840b6f81e96d6af3dfd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fa9745e5baf153228c865f201ac4bc942315cd0abcb99a35b7df8f8626e8839 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1011e94afdf80c63d73f9464ad8fbee197809e43 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e675979def9e8b8e202c2e316b6cbe30a5e6b6753de6748185774587ab820597 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b40dda89b3693ffa7c2f8325f10c6fc009bca819 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c8501fc29b3872c11c5d3a13f6e796e8836124c46404f06badf2cf35153e9350 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..029bd17a32a71647300df6268f29511105c1310b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2bdf4da901ce0ef093febe307db10bf9a07af4ff029b3fb611b399e72387807a +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..33bb1c35b361c87d79d21a84ec02fbff2ebc4ef8 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dae21593c99fa594aeb736e753c11c4798612ed715caedb20b02117caeb0691 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b6b3bba44368f159e800d7fdcef72f6aaca74b71 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/25/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59f4074c40cc778d02849ba631d0e3c811deb2944f1206b7d5b96c8ffb8e2cc6 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b6ce5d2ecf5f0b21191ecf14cd6d4f765c22a9be --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:196d903573981a30b374bdfaacd83cb14a0bcacf22727f7ba05968eaf27aa409 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..daa488fb6245ec1e22760fcfd20afb1028d16851 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f522f9aa816b2f1248dbd086c664cd7a3654be64c547e958796e8312cf375995 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2127ba115315b084eae1a37413337177c7c7d3a3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b76680c1d5fe86f8df0eb00aafecb5bcee11236e1a6bffe2a83b66460db038f +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7781bdd7c7cf95b91b6e5bcadd1483f93cacd3eb --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45f05cb2ff5764af400ed81af484ccec99f0bca26501bcb9f2c7da372da8145b +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b9e9488b281e6f700fe0cef9642099464e1af86d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8eb381138e5fc35788c9d3326b25e786ec53ab131f7755cbbad9e22d60bb2ac1 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0ffa9efaed2ebb94fecbfa7ff8a4f06635ea9fa0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/26/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:63ec21234a717775d053b29118490129a74c4b63e8ba34885e3f6ebd59339d95 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..076023ca2679efd7c323cd8623f311ea29e7041d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ed74f08eccd0b6c0765345135d920c8d0efa0053ad754077493b4929e3e7ff5d +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8edfb6501139dfcb8c4212f4ad3e7e8f09ec577a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:352b7ff97751940970dc5b289a7308fb3eef407a16a809411d57b9cdee38f173 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9fea5c72cbf6d5957b9ce89d1f68a2cccb1028f4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b5cb79ab5d4f23ff20b298e0f22f524cec68166a2d6ae2f9d8c669f23428ba40 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8f7fd7055b4dfde0009b4b6c0802d73b0bead73b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:41c7d864889fba5e30ddc2ce939828a8d8e057afa92ef39f8a55fd8416726c41 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..db7e9cebc6fe1ae187ce52b6389906a07b0de345 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2077da0fb080c06f483a68216e594d724e0336176c16b209775dbc8c6be3c281 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7839e8e0a43d71837d6f40928acc48deb6486cc3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/27/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:09c0a8341e7fe1a5eff17fa1ed69610b4a0b6a787eb1f21c0f79d2b6824d22d2 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e5dd96393cf361369e11925c8f648203cbf8b2e3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00abc2f171681e33c41538f0e817e8e410a143a3b07e05a8491438b804b07631 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..751dba9ae64f818eb8f6c1d46e9fde2a4489e252 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16cb7b60f71e1ba13fd88dac2e116788965237884dc2cc6ed1dfbdc7b0e069d1 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c9ebcd7553497f968eebb4cde25e3e3da90b7996 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0215109005b12a21af0096ef0c8cd730c8ad409e24a4a8b8f210cfb8c2013247 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..306aea7efbb24ed4df1fd3e6cfdf5b8a3cc458f9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b71cd43b84c1626e5bb0f8a4b97ccd4f92e2d5a34c88afb6fbfd9410c745a56 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a143c0d5252bd586caa95fd10853eda4e4f294b9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a368e35b0b8f4ba0773ccde9ee7783552e8c2923c74ac15f61e8e623b2873603 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..697ebf1df17085967a535646d7458003892c0a99 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/3/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d95e89d0cf7a1e5a60c49845c9258386eecedec0b4c6dc561ab1f876883de53 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..67ecbc59d6b2e282be7745f65e3c1291fd4f2e00 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a75ee3dfb24857122c71acc24076d93ce56085be863fce639d8555bfc0f2a95f +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e7d6a3322b5ea2938898120b40bca53b4d7ca84e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd81668f62d0d02792f2c56d0567110a4f54aaec6700da23db0be72efcb65fcf +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d7639711a318fd143be995594204af425518764f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c9f10911c4ff124529be1762d3f3e607ded0ca94131573c024616fd698f7c0d +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1f7786fe598e186f5fc2ae00c2199b3a770c87ca --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4f28f331ca474e3a4f3ed1189814c1eb57342c70ac12c1d46dc7413e1a87827 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..241268cf98526876debf885245bc1b2bed618006 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c745c3582233af3250dd27840a6bc5e81e0163ff24c743075706b2067171d83 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a74c8f46358e932716948bcd570b9e055a377c92 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/4/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:908f13773487ac95ef12cbc18ccdd0220e1e24016755dd230eae43c2387c1246 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d07e05ef522232cbda8f2f1fa42b9fbdef90b522 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5ef3a879c05db3f8b6f2104d36f62e8e9d21ebbd2e2edc4b3e11be6c3927da3 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..849cf8ae9a6ef26a6d9600b0beb221594a4cbf54 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fba825e7226a33c0bc221b34da85430c7ab3f0eab7732fbbc7bb00291c7437b8 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..74a8bfef3e89118da90213ef269111ca92425c0e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c46e40953a3af93a4602c34ee500a39ad2e7f8e571dcb8768c40ed70b2d624f2 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..05469e48fc905f61257982b7544b70fc423e9a52 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e148001f38aac876245973a738ab35ee8146b09c64d2625b29c2ae6c069c921 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1d4521d8d75fdfcd896a48d59424ad21ddb3571f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:38ba8031932f5184dbdc7ad07e8923bde76640bcba8c93a8a51dda3283bbcb3e +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7d55519bb04a9131420e1709526136f9548d2e29 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/5/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3d8c233629c5f9f3af981183e814ff068cd13edd61e040a1beab56478a5f0ed4 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..53e3b96c21386960c69a7b13157cf4a007d2866d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aafd9e376e22f9332cc8c8c5950ac257151c7b71efa9d8f670cb3ef9b08cc46b +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..667cc289e115804b4c1ebf3b58fd441cdc2e3d7f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0bea49acde61bb2b4c244524f09bea252a5d15ab60d0e847a28fb7a3552e7ab8 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3d4e10c69e4fbfbfa1d5917aef239cbf0a4116ef --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:872d795b537661907c8eace5b68c000db8bfe112b4035e37c0d2df697696a9af +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0460ca5c53b85c10cda6be6d1235ae27bdb2a5be --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a42a44d09ea7b779fefdfe7c4cb820c2f0ae970145ed4d3e892708996edd9fd +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..02f1a07a2221a4c0589dbdfb9a6fae9b6cce57b1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea5bc22e3d9a23e92dda2d50b45059141e6a5ee9f18f45111352f50964a2cbaa +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..92700d19e6c930a067b18e56d406cc7e3cac83b2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/6/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:031a1db3aacb78c2749775c7bac9258e404024d105d9e136e8280e58e4bf6f55 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b67220c4ee2acae5c3db2c10f63078a3955a7c37 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22da7297a5b14066a429c1d579739ba8c2d796503ea5b746d481f1e8b4c6e072 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a37d0c5ce41cfa8cf81abac9b2be32224a3ec4dd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e3dc12e243dd5852b849741275a753a38958d592200e7b8f0e9586657be5eb0 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fe589090254e3f35b1b535455ab310c13b786cc5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08050daa0bce01c808fa284b5a7c2b0ecea6d8adfc4122bf06cbffd43119b009 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8fb7ae44a7518ac3b4aa654b43f4fa52a0c4658b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f7a23b0cd4519ed6ab07c35c2cc66cb5adf04f7874b97d831ac902155b6cd18 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0b6841ea9bd15d6b2aa2d2778ca2ebce9cc7a0d9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:494ebbc5324540290e89fb0ca21887f07dc90f8a6b73856444ae6275fb4129b1 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..69368aa8be16e26f6bb2e055933ec078cad488ff --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/7/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:07d5283163f6b710363010166697600d587044c52d6829db373c23891026f4a2 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8c3bd0c2d3be9e3f8d574e3d1771cabfb07451f2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:473994b658e3fa74a0200e22e4db6c9389740c3e67e4d17c192f327b0e864e8f +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e8b2491cd3e78ed08bfc9418757593965a221e32 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:961deca816c80534931ccf908d8a73fd09414eda444e4c198273e4add6a342ea +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cccbde3654dcebb12d968b2cfd436b66368f467e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5837abcc314190094823c8f508b21a6382978aa27370d63cd7b99a29debc2e86 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..db61f1bcf9ef0cd48e6691fa75a19872103d6298 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:941020cb9d449813410cdf3ee5f43f2a954456c689c4734e0b240f564e909ae6 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0275297b9d278df0466946b2a6ce2efaa1866d03 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35ea35d2317c36537826a2bcfbcde225c1b5f3bcb23dba64927cd95e7366b857 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a65a769f692d8e4a9c145927bb30be4969354ee6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/8/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ec987c0373bab80ab7d9ee3c3b1b3fd958e794f36f5e9864f5138528fe81dec +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fb0939f14bafafde5a10ee0d4a48a5f7b691aba7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ffbf1091064c96af564c137ca667a41d197ffd472a3da2c811122130028e29a1 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..116da86013af30451c2687af6705b5485afd04ea --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9fdb3eebb9909f523eb87bb1e1051b81f6af1a37af28d6bc1179350e24f331bd +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1c838185aecf4d6772db6a462c5fab3c8bbd37fb --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5fd16c8838c3806c4fb433e4078bfa06d46978a20745b70735daad5c327c78f +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a6ebcadc98a7caf1e3d1cb3d35b03e6dda33c67e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68fdcab766e4aa7af9d8bf5aeab57c6ebc78d0210709c5f52ed2be69c051bcb2 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a7dc4efc7402da87848e1b0359849d6aee9f9cde --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f701d236d8c2305d240a366fd9bf3c84f7558cc5fd8d985c6798186b8afc0f18 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..65277fe19825bc44626b579a9b0d110086403290 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/decoder/9/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:380b44b88179755002c3b1720629ace84b1e35055dd293e4c49ec5c791d58c20 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/final_layer_norm/pp_block/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/final_layer_norm/pp_block/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b7e4d02794f3d1c97232646fa00f80bae3108770 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/final_layer_norm/pp_block/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:def30e25ee8bfa037e262c83604444f3fade94fe9ad6d13bf02322215fd50215 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/token_position_embeddings/pp_block/token_embedding/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/token_position_embeddings/pp_block/token_embedding/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8c2f3421466e1e2a603dc370744a1cf0ce495ea6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model/model/token_position_embeddings/pp_block/token_embedding/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1dc5adbb2f83b019fdd1de894f1db496251c7be046e055c71d492d82c09aefaf +size 788005112 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model_config.json b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6f54ecccc80be176a7a8f7536912a7b208c9fcf5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/model_config.json @@ -0,0 +1 @@ +{"bos_token_id": 128000, "eos_token_id": 128001, "hidden_act": "silu", "hidden_size": 3072, "initializer_range": 0.02, "intermediate_size": 8192, "is_llama_config": true, "max_position_embeddings": 4096, "num_attention_heads": 24, "num_hidden_layers": 28, "num_key_value_heads": 8, "pad_token_id": null, "pretraining_tp": 2, "rms_norm_eps": 1e-05, "rope_scaling": {"factor": 32.0, "high_freq_factor": 4.0, "low_freq_factor": 1.0, "original_max_position_embeddings": 8192, "rope_type": "llama3"}, "rope_theta": 500000.0, "rope_interleaved": false, "tie_word_embeddings": true, "use_cache": true, "vocab_size": 128256} \ No newline at end of file diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_config.json b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..264b0ec29e20f2ba890f058dbbf653ed4cd90210 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_config.json @@ -0,0 +1 @@ +{"type": "ZeroDistributedOptimizer", "parallelism": {"tp_size": "1", "dp_size": "2", "pp_size": "2", "expert_parallel_size": "1"}, "configs": {"param_name_to_dp_rank_offsets": {"model.token_position_embeddings.pp_block.token_embedding.weight": {"0": ["0", "197001216"], "1": ["197001216", "394002432"]}, "model.decoder.0.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.0.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.0.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.0.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.0.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.0.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.1.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.1.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.1.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.1.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.1.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.1.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.2.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.2.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.2.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.2.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.2.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.2.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.3.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.3.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.3.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.3.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.3.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.3.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.4.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.4.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.4.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.4.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.4.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.4.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.5.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.5.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.5.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.5.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.5.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.5.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.6.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.6.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.6.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.6.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.6.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.6.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.7.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.7.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.7.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.7.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.7.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.7.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.8.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.8.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.8.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.8.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.8.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.8.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.9.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.9.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.9.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.9.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.9.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.9.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.10.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.10.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.10.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.10.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.10.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.10.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.11.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.11.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.11.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.11.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.11.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.11.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.12.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.12.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.12.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.12.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.12.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.12.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.13.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.13.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.13.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.13.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.13.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.13.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.14.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.14.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.14.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.14.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.14.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.14.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.15.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.15.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.15.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.15.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.15.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.15.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}}, "orig_param_shapes": {"model.token_position_embeddings.pp_block.token_embedding.weight": ["128256", "3072"], "model.decoder.0.pp_block.input_layernorm.weight": ["3072"], "model.decoder.0.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.0.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.0.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.0.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.0.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.1.pp_block.input_layernorm.weight": ["3072"], "model.decoder.1.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.1.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.1.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.1.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.1.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.2.pp_block.input_layernorm.weight": ["3072"], "model.decoder.2.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.2.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.2.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.2.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.2.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.3.pp_block.input_layernorm.weight": ["3072"], "model.decoder.3.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.3.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.3.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.3.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.3.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.4.pp_block.input_layernorm.weight": ["3072"], "model.decoder.4.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.4.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.4.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.4.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.4.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.5.pp_block.input_layernorm.weight": ["3072"], "model.decoder.5.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.5.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.5.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.5.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.5.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.6.pp_block.input_layernorm.weight": ["3072"], "model.decoder.6.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.6.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.6.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.6.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.6.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.7.pp_block.input_layernorm.weight": ["3072"], "model.decoder.7.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.7.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.7.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.7.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.7.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.8.pp_block.input_layernorm.weight": ["3072"], "model.decoder.8.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.8.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.8.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.8.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.8.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.9.pp_block.input_layernorm.weight": ["3072"], "model.decoder.9.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.9.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.9.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.9.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.9.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.10.pp_block.input_layernorm.weight": ["3072"], "model.decoder.10.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.10.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.10.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.10.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.10.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.11.pp_block.input_layernorm.weight": ["3072"], "model.decoder.11.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.11.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.11.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.11.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.11.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.12.pp_block.input_layernorm.weight": ["3072"], "model.decoder.12.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.12.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.12.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.12.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.12.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.13.pp_block.input_layernorm.weight": ["3072"], "model.decoder.13.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.13.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.13.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.13.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.13.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.14.pp_block.input_layernorm.weight": ["3072"], "model.decoder.14.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.14.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.14.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.14.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.14.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.15.pp_block.input_layernorm.weight": ["3072"], "model.decoder.15.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.15.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.15.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.15.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.15.pp_block.mlp.down_proj.weight": ["3072", "8192"]}}} \ No newline at end of file diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-0-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-0-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..411e9265cc52d0661b3bc57e9a1145e54e164155 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-0-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c987591964db4c90c16011853712707952290579bc486569dc58347dc380d5d +size 12028411794 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-0-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-0-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..20758a9f4037c28c5facc098d7458b3a30273a11 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-0-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e336740e19cb74bf133b3504ae6c74bbb9c5f6cab466afb4452b8ae406a2e81d +size 12028411794 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-1-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-1-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..cbc11e6e0978f5ecc7d5f80207e45c498fa79aa2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-1-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:739796537e860c9be4d0338e9e22ca1ad1e94115ad0ae35d491008b00fb69364 +size 9612328561 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-1-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-1-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..bc76433dfc5e59370d3431298c72fe0080773b0c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/optimizer/optimizer_pp-1-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:81340fbe2a59f2586c30c05eb3c4b82c481358e0ba3e488b77fd81f06654b7b8 +size 9612328561 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-0-of-2_pp-0-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-0-of-2_pp-0-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..dc0a783730805ef41d420c8f03d8d549604c13c9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-0-of-2_pp-0-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d81c15fba15f9a607a4736986869dc27a504e4b6e09a37603307cc1ca5aa3a52 +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-0-of-2_pp-1-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-0-of-2_pp-1-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..0de2e4a921c365e895e3b5d2820bc98c65674113 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-0-of-2_pp-1-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b19db2d5d750b0c90eae9b28695c7e22e067f62104274e6aceb33e5d8ef6ae4 +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-1-of-2_pp-0-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-1-of-2_pp-0-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c839d319130523d9bc76069b614d153897b0f2f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-1-of-2_pp-0-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3827577ffa796c77f14fb1f3dbe7a17cff3844ab6a99a7f6240a1f2ff2b155c +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-1-of-2_pp-1-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-1-of-2_pp-1-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..c4c23e4d860e23af92b26eb02a6d95d2f4784d05 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/1000/random/tp-0-of-1_dp-1-of-2_pp-1-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:900ce55cda2c6462721a7b34177d57d42e1ca1170f7f99128c90c0d1395a272f +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/checkpoint_metadata.json b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/checkpoint_metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..ad79e65e82207d3adeb4d4e4b82ced9e1bf38c03 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/checkpoint_metadata.json @@ -0,0 +1,18 @@ +{ + "custom_metas": null, + "dp": 2, + "metas": { + "consumed_train_samples": 1024000, + "data_stages": [ + { + "consumed_train_samples": 1024000, + "name": "stable phase", + "start_training_step": 1 + } + ], + "last_stage_idx": 0, + "last_train_step": 2000 + }, + "tp": 1, + "version": "1.4" +} \ No newline at end of file diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/config.yaml b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/config.yaml new file mode 100644 index 0000000000000000000000000000000000000000..c510f7ed89517e06ec040f3d42cc50a84ee2ae7f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/config.yaml @@ -0,0 +1,107 @@ +checkpoints: + checkpoint_interval: 1000 + checkpoints_path: checkpoints/llama-3.2-3B-finemath-4plus-finetune-bs512-60B-local + checkpoints_path_is_shared_file_system: false + load_lr_scheduler: false + load_optimizer: false + resume_checkpoint_path: checkpoints/nanotron_pretrained_checkpoints/Nanotron-Llama-3.2-3B + save_final_state: true + save_initial_state: false +data_stages: +- data: + dataset: + dataset_folder: + - finemath/tokenization/finemath-4-plus-p0.0 + dataset_weights: null + num_loading_workers: 0 + seed: 8 + name: stable phase + start_training_step: 1 +general: + benchmark_csv_path: null + consumed_train_samples: 1024000 + ignore_sanity_checks: true + project: llama3-3B-CPT + run: finemath-4plus + seed: 6 + step: 2000 +lighteval: null +logging: + iteration_step_info_interval: 1 + log_level: info + log_level_replica: info +model: + ddp_bucket_cap_mb: 25 + dtype: bfloat16 + init_method: + std: 0.041666666666666664 + make_vocab_size_divisible_by: 1 + model_config: + bos_token_id: 128000 + eos_token_id: 128001 + hidden_act: silu + hidden_size: 3072 + initializer_range: 0.02 + intermediate_size: 8192 + is_llama_config: true + max_position_embeddings: 4096 + num_attention_heads: 24 + num_hidden_layers: 28 + num_key_value_heads: 8 + pad_token_id: null + pretraining_tp: 2 + rms_norm_eps: 1.0e-05 + rope_interleaved: false + rope_scaling: + factor: 32.0 + high_freq_factor: 4.0 + low_freq_factor: 1.0 + original_max_position_embeddings: 8192 + rope_type: llama3 + rope_theta: 500000.0 + tie_word_embeddings: true + use_cache: true + vocab_size: 128256 +optimizer: + accumulate_grad_in_fp32: true + clip_grad: 1.0 + learning_rate_scheduler: + learning_rate: 5.0e-05 + lr_decay_starting_step: 50000 + lr_decay_steps: 10000 + lr_decay_style: linear + lr_warmup_steps: 1000 + lr_warmup_style: linear + min_decay_lr: 0 + optimizer_factory: + adam_beta1: 0.9 + adam_beta2: 0.95 + adam_eps: 1.0e-08 + name: adamW + torch_adam_is_fused: true + weight_decay: 0.01 + zero_stage: 1 +parallelism: + dp: 2 + expert_parallel_size: 1 + pp: 2 + pp_engine: 1f1b + recompute_layer: true + tp: 1 + tp_linear_async_communication: true + tp_mode: REDUCE_SCATTER + tp_recompute_allgather: true +profiler: null +s3_upload: null +tokenizer: + tokenizer_max_length: null + tokenizer_name_or_path: meta-llama/Llama-3.2-3B + tokenizer_revision: null +tokens: + batch_accumulation_per_replica: 32 + limit_test_batches: 0 + limit_val_batches: 0 + micro_batch_size: 8 + sequence_length: 4096 + train_steps: 28612 + val_check_interval: -1 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/lr_scheduler/lr_scheduler_pp-0-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/lr_scheduler/lr_scheduler_pp-0-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..72269804349c8cb4762c5ca159b3e4db36c735ae --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/lr_scheduler/lr_scheduler_pp-0-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d15eab5443380661f6060e52f1bac0f11c98c689f99fc53acf5e95059741ebae +size 3248 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/lr_scheduler/lr_scheduler_pp-1-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/lr_scheduler/lr_scheduler_pp-1-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..a5d054217ca67d7e626da90b81046804e0ff6ba4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/lr_scheduler/lr_scheduler_pp-1-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbaa202e60a75b58e94f93135cb29e0387adb214aa5748b7e9e265509f2ce40 +size 2800 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d9a9118f7edb00992753bf3af32f85a7a9449ea9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac5f8eddddbb3ace6ea7246ccbe018c198bbdb43279b5e54c836c82885ebe2d2 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..36b94ffa6bfb1a0587e46dad4f0d82553fc081dd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e48a95df7b7aace39f4483a8c0a3288d2d4ff78bcca4a612392bd5a486d36ea +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..701d3d945b0c82abd5b48dc332182a5b76072ec7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a40829638f6378293088296c14618985c8911445a23c8613e499474546fcb78b +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0530a91395226a67186488e603287c8d3dc408d0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f5f3d365018528df86e2b04e3879d19ab552b37b2f1e5c4a3aeaf39c18f712e +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1e12d242c7dde9492711685f97c26f146a155f26 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aa9937605c804c59e5ae14d3c4f8362ca67233ba6f19d4f3b1b5f632dc3e3741 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..72093a686bc513f567334658fa928622b7d70596 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/0/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70f93ebd29c007196b204f1fd9f0f9eb0fc726068284fcb0ffde5148e96339f1 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..46c7ee0696db16e157189557da74158454039d20 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92f3feee003708a47c5271ea4d1672eb50108dcc6f838598d8a2885826047525 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e02e7886fde9344b3e515d3919dc973e7eb69b1a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a8f2ea0b9e684f381d5d84a54242262ff59946ad400f8e1998bae1850a856b3 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..07386c89f5a4dc4a15f450d09790cd06ed5bb156 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0296e8861190d94ffa1a7e30b0711aab8750b434399d27dade3318f08f460ba2 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..57121dfa850cb856f367c9aa4a1522ad2e3f5d54 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d0f58b6be3e74f924c90eccf41c0f19691382ab76168f92642dbb913225b9541 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..97fad759b7c4a42087f45f3196c3cd9e5e0ff8b4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:458c9a464cf9eb909a346ba11782f4b3aa8f8d3f3a4df42cef4cdb2d34a26977 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..810276497d66cfd9bb8ebc852534aa16801e03b1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/1/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3bb6ecbff703e44184d7de5966b66b8194c592801f927749e8188820b75fbb98 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8496eed02542ca2171c0095d8e3870dbc25e95c1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a20db4d69f36d4f138003f4d57074cf94ff246f16076e58b6e81d09b8142b07 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a7f324b983ac52eff4af3e56fc2c8008292ff968 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:43f43ecf19bc161ba825f654b37104749e622f24fe6c8d6571a72f30578d934e +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..86177e86779cc8fc3c8f27981e759ab2b79aa8ca --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f94442abad8ba9cf05482a5493053019a73daaaf88417f1807640f85bdc8b828 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8a56a4fc39d5366cebecc194fef6c95a02e66d1f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe75dc07a78850ee9e5b0caf96f469e04e240f45e8385b87b78858fb991ae877 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b04753799b5c22aceb9e95c2826c83b3ad2248c5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c38b143773f484ac84108e3649c645b5607620c244e9e43fd0d7dae8be1762e +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fd0860a7ccb88bd486a374a45af463cf7f8c62fa --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/10/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e5b14ce0982d3b7124ff7c632b651dafcbc30ef2df4dc415a86574dffb2074a +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e8d886579ae275272d741302b03e551d668bd90f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cd2182ed8eb9baa39fc65bd61be6bc791153966a0b8802a2bb288ee87102766e +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1c4b41d4a72bcf6b53bae281dd3e29d445e6ce0b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c93519e8c2baa34d674997e649f0019335e6b28b65a50bc46f6d2a97b26b27a9 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..35907897a03736b663a23e2cc30a3405f712153e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:41708df61d9cad4a80b2795c84bb7280df8285fb5a15ce6b36a5bb46e4b53410 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6c0504c8ab06ba9b92d7f6d35f3261bce24f9ed2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a226593de8e98271b76dff458bafc2befc77a31901d022c66a177416df2f4416 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b19a7a2f5b1b0ba0e6b843013e42a4caeccc0e44 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a17905e1a8215a04c5c4019a688468d6abec324e32c3b26cd6c3882dbdeadfe8 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..38efb0b8dacac5a9b70aee3f354b4129a4d29b6c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/11/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:37560c1d09cc9275537a13abd2a682ed2d065a627edcc2fc4a951db96efaac52 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0e574adc97a50eaac31194b090e033146da23f0d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:67615472725621c191ed82d4e0bcee5df2349c9c8345b87ac5ed6e719d7a07ed +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..de98fbe01fa84aab5959240e8d6f4e7cc30597b5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:30b6695420f887001818516673d2a69093ab718b046fe5efc6aacb613c6b254b +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..65a64c109362726e1ba390d4bb7f53c3fa360df1 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45fb2172ab358ffa9229fb23c7bd2af297f78256c483b50ce71b145742b4a8b8 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bc1ad0b030609c205a33c398fe7dd7429ed483ec --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6480c3e4a1bf8ddfcb6929a83423857bba7062c565f73b02f5d56ab285d2b855 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..336743c033d7f120f077f2969b07b5a82e3b1180 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90c5693c211f054a8735bf70b89275066bf89b99e8976b5c8b55ebc359a8af3e +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..079f3f5f5beb73377547685cd7cfc1403009544b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/12/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3d8cd7048903216b3036506aec0d6eeb96330a5eb184d607d85aec7046886638 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..808859e4d3345ffc8e74912bfe2a7ca19fa5f4a2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e482857ff2e3354d5dbf385277347d8bc9267e497e01186ac9f292df26d98f4f +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f69b0686fc774c332f56d7e29d3e171ea9f0382b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:303353de3ef21ccff9bb899f42ddfcfd23e3230611c970ae660cda69c95bdff7 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..24ebc389bdb6ed8b5aa4768e9b2a7a3727d63129 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4141bff60e99735b2873e7ea34dee1d99bf197ca734fbd5735f8282eb69e5483 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..36fa62be7da9fcd8b3a4be520f1131416981a892 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51b774b1e2137cbf587e48b714687883d4ee6b8111b0ab2fc0bb1ccc346b899b +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6218d2a8d12f9e2e28394a4a27963beefe734f54 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0da97ed0b2bc845e1e0c28d5eb7c568c62deecd2ecc387ba55e6cfda118838d2 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..18ccfb6513c843309222a598d5b1e7cae9a0cd53 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/13/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f0916edcf3971f870408c323d218380a5a6e3cafba26c0dda8c8c81f745a1f9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6dfcd63f644afa2e0142dba968abcc2d5bb71117 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3e51c6bb90eeea02804454210bb47ca84fd8b75ce91e9ca0089317ddb9049cc5 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..350581635a6745a1d5f119c99fdf8b7afe6ee7bf --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc1b8a58744a307f244eb7c7f2e37badef0395f90b8636688f11af255a773cd1 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f82082b12051bf3f37d7eddcec069a6c1c12f30f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c6169f940a33c6e142057ee2a63e61039385a4ffb46a4eee00616ddb07713c1 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0ee680fa1d44189164da4e80504190f4a981fcd7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d715aee32685e7677d337acd2225b747ea7762cfdf043b91cdb6716baa8961c +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..598376d3bdbbff1125fc887968ec6d480fe921c7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:420eb10e4de4ac88b00e1b0eb7b63d005f59d5b3ead061dfa9bcd513085e5d6c +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d6f07c8b078a6df5444f2629dc854543a2c700b5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/14/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7cf0d7fa6ef3e450a4054deaafd9da85051618276867b900ff048735bcadfc3 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2083f9d6bebb14082b0c5a47d122f9f3747927ef --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:992179f872b3653222a7e932db45265c64da7760d5bcb7d64500044212467370 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d68fe98757d93b87f1527190cc7c4c2198d9967 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a0d10b7b4ffd7cfd03087b5e1727fcdff1982663c5c0aa24ba6954455962785 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d13182d71659aa493039a14973f5b8b484e6fcc5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d0badba604e8c0d747e87670129c53c947ee71ef4f9ba4452673606952decfa9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b8fb0bff3105dd80192e4810c1f702c126ffec42 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:897f0f581ae91a13027a999af0e0b58be0f6331cafd4a9b186808c411208c8e2 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ab13a5cba1ab66308ef19ba1026757b888eb067c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b322f35e79fb7592567522f7135a910b0a1c2f905ff7457468d560ce739ca1b +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7e8e7ae6717da8914742eff6aec324d7cf8cb07a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/15/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c2dbb3b0826cdaa1790d3757ef93559615f492a92c21fedee0d6f6b3686f81a +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..19a05a8133e1cdcbfd1ee0cc5591121fc67c64a5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5dae2d1bd1a47b36610c68dde71dc6ba61447448813fdf3f07a4ad669620035 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0735ed03f5c20bc68a3042990ea69d5b0e8de636 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb6cf4e958e6780ac5110ba02a76344e06132d140a5f969e18433874fb3820d8 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c2e4bdeec776e72ed01bf9f7bb391a05c29c6387 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f6b1f99d20a62728b4f429898cc00cd7272f84ed5ec90caede1f157403ef3c1 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5d95783e52faca4e8e7acc591a7cdaa1818d5afa --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:daf58e7e4f63cde3d2d07ccf70df9dec80c9fc7546f8fd24d5676c1a8bdb78cf +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e1de6676cf1c602c34a70da43f5bdd319c157390 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e7fba4cae69a8376c85919699354c7e6fedacd67c67ede66d96c7ea7372f92f +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..737189099c2fbf0c0d44d2f3607f3403bae1a33b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/16/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cba9e1cbf988bfdc42a7f0bb5abeb0745e050ea0d197b710f33e6976d6acfefc +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1e76f1c2346ecb886833e72389cc09a809979c30 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:228289fb83cd118856a0ee25fdea65afcb84ebdb64e872d55519ebd3bedcfa9c +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..76b73e330211f2f3bf803b4739e9e4ef05abe711 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:edbb3908a9f527b40764693cd0c82d9c788fd1d32c4c43d4a9783e202963d5f1 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8516811289c19bc5736c427389248246e6033760 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1205e6eae6e0427554c7e77e0f8ace5a5f72429446b25157adeba5be8b780791 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4b5009c6653a068013760d957ea8c3128fcfa13f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ffb7babb2244bf91f1aab8e7d6e1d3582ae60b56f64ac9a6a32cb83773829acd +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0364e34495a71a5697173ef21530bab7bc521f4e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d94f760edf485bcce8ca8ebdcf12f8703ac2fc352e033e8b6a74452b3d0f8010 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e1e816337be3acc02cf5dadfcc13b801fa73b03b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/17/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b7c04946b141e72d66c8429c29f95026fcc18f1527911239ec67c5213420de9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c2aed7adaf6969b74505ef3cbc339666c096d853 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6c6cf688af00e37e97ad679b0778df08efd51c2da3105a6a4d897d282a38e0e2 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1a8047a8e7539a1dd966bda6bffc5b037d2736bc --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9909fa5b945ead3781b7e43fef0dde8b3e0aa12314833ded354a54da5ad218fc +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0d5f73190a0bdcdc9e9a60cd6f439ce60064efa6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cea666030f9ef598b8a3efa07e9b26fc00dcc333c30ca59de2e71f794349af07 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..dd1c79a681f524c958c96e1b2fa6ccb54d8e1262 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6057b00222559a858efcf4528b06bb74f96617c2614fa2871ee9a0ad87d53e10 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3732a6ca3c337f1313676940fa95d0973407edb2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:531e8a409427fa9616fddd2b068df62f52767b11d03ce2be7a7cabbf117208ca +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4bcd4c76e8b603441693e8a731b292a64bbd56a3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/18/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:85fa556f16a05f7bc2e7ca2012bc6e1bf1db1a9e2369c792a24a2029273527a1 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..674ad7ad354f7b3a039c1b34645f14e8d325b191 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cc82f89664de9613148d20c181174ec440c76003007c7bbd7182a68f8b8b9432 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1dc69df3aa6698486ea88c716ea6d930fa29218f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a112002badfb42a146aa27d4bcb9825dfeaae81bf040be97e565d6bb81b1724 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a9889449cefd2882aa2222fa5aedef13ccc6d5d3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef0604122061551006eea39727e3d5572c5732c6315560782743815fd0c6f374 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..91a1e187e9106928f348e5b071f9746583e9c381 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60300ef435ca6b167da4c2ad7c155a16aa0a567b830f8bd6bf337d36aadc0cc9 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..10b285101164bed8960ff91b7a73eb3271bc2a77 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec7fa262c2e2136a37cc6ce234fb76586709e886adfe98d031bb218f8df59e74 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0bbac6b78a690a9ce727ecdeddfcbbb4fee4a8aa --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/19/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2ffa0cc65fc4e835e8741d9899e984b38883eb74d52f64641805115c1ab5cb3 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e304c27ff745181c441ed59bc112ac28c7513f72 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:25f5366ddd14ae9cd026ed6b9972900a59074ecfd82c8b40f6f443f5968bc363 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ba5407a222cf4de711644feaf4c2a8810aa3a221 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aea73009b0f35b0b673f41e8667f89aa31e6bdade67191f93eaefd72a6f3965c +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0a139e77d7e7fb33a4b3f98cd4a1f38da977a53d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49298f39dd84ab87876eedaf6b9fcb7cc2d0a097141ba498ecbc09b48d67f449 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5c0650d9d94f467c78ede5141726174a8bb16b5d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f7f8d0f4a052ddb759c5cf2eb63d79900d3831128b8a2ff7c1b7f8763fe6e91 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9e80a4da1771ae3568d07d3de9864461db273f29 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5adf498b53787283815f9e96c9431a93f5a07bbba2a8bb818318a94f5f1831fb +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bfc29297cf2acd0bec94e0cd2e1e17821620f73f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/2/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be4fcd183dcd6a310fd129c7e243c23e8fbaec06d119bfa8c3f1327bf8a37744 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..045c0afcecd55d90fba91d889de35c0af1428bbc --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ba87e4f3089293e2fc57319f86bf33f676677d4b24d52a8def2e240403d0edf2 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4479364a3efe714f726d8991ff1917f64b8d983f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0607a47bb2b44673541acf20ef3facbb16aee5b1bdf6709d7f08f1b50a43f66c +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..735ae751ce27cd37f349f4b9e62396694bb553a7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eba3c2500d70ce796ad40be47fb626584e5b66b7ad10543aa90e882cce52c1a3 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5ea38d9c319e127557d7b9c9bf1394c2dac7ff60 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b7514369272a13b5809755a2a024305ce6fa480eee6ce12fdbb73ac4a9234c0 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b7f85dcac9c4c074091c57d12f55b092f36676d9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f48a16e50310877e5a26017ad36cc17973a64096a5589f05d6bd821c29cbd3f7 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..190412bf7b0b4791fc9c1add5f4392bf3c1b8408 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/20/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3857ab1ff7932056f72f15ca5ccc203b4247684796e4682875b6b9e6cc913348 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ed1a741ef6017384cacd8449056b95f5539c6b43 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfbec277ccb101e69247b6c5b0e675bf8cd19e005ffe4755c97878f77fef3794 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4ca3b0310436ca9c08904896671ba213f78692a4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76af0308524511606c1fdf7b1649ae023d2977360e19ec4cabfd461077ee901e +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2cc5096f0c184d6ecf66c2b435cc8553bb89e1fe --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b20f8c9e7fecc8e5be25c6da56a2b91515a55b69d23027354c6cc1b295c3879e +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..adbee34c0d7bc6f26d1475ab30fe5ef327142e26 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f70106f16e0903d7997f3f24dea6ee87ffd75a468032687677fa3c470a43428a +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..88bd187a2abed298362eea2d504da26b9dba0ec2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f6ecda82e389b5e76aabb071c2a58848cc2be66bc44000922332392ff9bcb94 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0e31d7d8119c7812c11803823c352be5e53b6c94 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/21/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d9495ff043f5aaefbcffa353702ac02acab133acf504628686f01d7cc007dc57 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..972fb2921ca177c9b4c300b1124e8ec8fbbb468b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9b41be70e37d5aa8ae2aa9e2ebb6d5318732afbc702d697dd5bae7ef5733552 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5034cd2921b652c04212022f39d1b543a195a6b7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7e606f0bc8e06ee11f37c28f4d3d81ede39332a6635519fa22b0829c0a4ad2b4 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2aba1edc4f6f7691f675ffaec6d7dad8b18195c3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efd8f2cff93e53b09e1995fa45b87d41e33e507ce32e93c7ed8104dd09fa5660 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3e6639aed3cbe8bbe6e439f8e5405c89d01bb602 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b02472fce54cdfbf23b8c05a47749a06604a0c4f26df27dda2301fc41b45356e +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e91e0aea1c3cef0b00dc3dbf39e42da128f30ca6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:63c41847b4dbf126c90bbb8acd3635fc1b03c298b9922cd8b10e7879f2f3c6ed +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0aa73b686fe37bfc5242ca07d52c1edf6569c0e0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/22/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:29d0dca8019edb4497471f32b89688263d506c1f94d1083b8d16bce7f2b63f44 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..88608084b3c0526aa890dd4e529322707d9101b5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e00eb3f697b279d293e8db1ad13afdbe7cb660b5ddb98c9e151f4bae32bc7bbf +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ae1ecba1ec7e261805ab92ebd7a9d596d3d9c209 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f65c3690ee3befa37f4db320cb16131b621b6666204758e77080912c059b86bd +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..68403eaa7dd7fd5876b6240d736f75f5c656ffdb --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01ad69b90d3ba69c9d5467f2efbee563ff2e342446aeec6914b959649e0e1cde +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bb0c864312d13fd16a4a2b26da6b1483738d860f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2826b2c27b311b869b332c6dd96469918a04d2067e7dd39eb8bde90a91008742 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b7139f0274ec75fa7ed7bb34810fbbb842b739d5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:385c07d523634f8e0d5060151d7b5751cfc80e360eaf14df337e97c65a3c4507 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e8b07b41099aecc1d31965f5fd57c14987212a63 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/23/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:75aaa445f679ec7721d4da7e3fbfca4521f6824b3813bf4bcef9d50a73a1ef31 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ce329cbbe2f49810c096de09c5166d91d6ff3506 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:537880f15cf7190c7113523ebffcf24c8a43298347b4f3eda2d5fe7713f10003 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..08837525a77af370d3733d2893919bfc09599c3e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ebebefa16fe6f155f964655803f7b5b879297e9bd140c1200951223f88b0da12 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..afdd0fa072be78492ba35b740333c626ac1c232e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76cd765a34eca9273e25753ac2c62f886f998d565d824811538ebb551f72bdaf +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1eea59658fb6077a3d65f99770be560e579dffcd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:54ab759456adba658d278ed41390e3d821cd6a0b05b5e3dd339d455d27564edd +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0ed73af0431955aae724dc52963afb9e325af123 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dad6f6096ec9e6fa7d1d397083be2ae1a78da4252504b2c28d79a98437c09008 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5220eb699f32a68857f23746e88c159a5f74d2c6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/24/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb11405c8d4b7f4ee548d19ecea617a730d96eeff774d43da515007cebaae85d +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..55ee33bbff404e95dea6934da6a9cea12d715a1a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47a6901f1d1d47f3c442a191c17872b5481184bafe16e1945eb9767d2f78c20d +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7d99711a8c1bff140cf2e92c7bada2ad7cf9b15a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf7ba47797c7014c9d782395d0dbd2e686ba44b56409955eb713df57dd7cbb3d +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1275373141246ff1d2f697c598ec5330651b3c4c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6203ce87c935d2d9124e6fcf007b426f40226f840a6a9c66fa2eb699a48938e5 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..451d93c1ebf6eebccf510f5a89a34a02d3fea7bd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:624a50b6d7bcbbf2ea09a9a941268c5b41edc281452c04d48fbca7c5df07abae +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2ec88de6e94c376b602f1678f0639ec201b96ea6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70a1feb191f25dc35a38d31a355c78c5aafda3f949fd55f222edaa8c18bb112f +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..454c20d09c42bd71f18ead5ee9387f33f1d3784f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/25/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05d9a9debe06fd569f65e175c4b12460e24e7eb2c50469bdb8c3c6e408036919 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d1b5b4f76f3a8b750e3e8d723e458d1f6a3cf9da --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8839739519df063807293eb188a72403a706476e7cf02a4eb8f19591f0c6cd2c +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..34fd5b9f585f1f2a9747e5304782cac1353e291a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5936222fc6371d2e77a232aa2bb1f49b0abf796d559dbb91decba94fe767fae +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3ca40b3d138e302b936cf23522350790ed4062d0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:faf4e4d0c454e59a7ca558e2ab0e8e1a5e6e683a41659213ccae030d87929e04 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e0547f26aa0df05acbec4dc9438d8113fce302c7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:201be38192ac05685dfdec08c55602d45614e96daa84b3f788f02840320cf7bb +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..27743e14973335dbd0d752bd4d697be82c33c7da --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06c58f22e56d8bbeab4e7684f8a5900a5d4ceeb2fa0d6b5c8de6ef3e562a849d +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7dd0d0f209bfaf8bbf19720bf95bfc79bd787aaf --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/26/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:670118772fcbaf43bbe059be7147dd843771d1fd5cbf4e8f3e14f165b587e43e +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5f1f1b60ff29df078f82dd1bc738004deb364502 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/attn/o_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77097ef196016322fe1bd0585a5a4e9d6b3bf582c9b40262375aa7e337c8cdbb +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ceddebb426da0ec661bbf239ac7e8c52a3b47424 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/attn/qkv_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92a939f82236bfe17297448d0aa3f8d327c02f446f8f2f73ac24e1e75904dc29 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..01e77e134eefa7de7ba7578dc499c1df94083705 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:686f625ae09a40a42994768f62a2fcd40a384e30f229012d5422e835c3306291 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5e3e61b38f3f36928a0e07f32d6f92fe8f7d398b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/mlp/down_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4cdedc2e99065787da0528de0c1121c66a880e260f0b9eb0e8565787bd3560b8 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..329c18339694575a04541ab73624fa9333d55853 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/mlp/gate_up_proj/model_weight_pp-rank-1-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9aaf38a826481b7d4607d1de82d2e5a69111450cad84cb4c0cb4c9c8b316d61 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c622c1d6f54186f0061f826e1a8e0778f94516b5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/27/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:073511327af691637967d452c7ee31ce148d2e6512a2973421b2335db31f4ea6 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a4f3e217af67c01b1277ad4f8c4c3893244ddea5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3eb86ebdc2e49a582b322a52175b10618bfcbb42b030566c40ae64851f5406b6 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e071e91e5f6ca275c83a2f6bbc3b9e33ee9c2e84 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e38ac66332ee923ac22fdcf484c8452bced974fe6bcb93847fefb4a768416767 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e503a9930baca1f78e74e66ece1faee759144944 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e61bf79bd80edde091af217499b9a42edbf1937b8ea8b77f10701ef8906dd3a8 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..272ae54cd7856705ef571eb981e348528a33210f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:394dc087c5876db2b0176f73f70b3d00b5aa7cd044fbda1f1991347314cdb4d8 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8642ac23923a16791a4c8f374666a4c2c6012444 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:67f0339be60000c8ac05d9ce14cd95f6df6891bffc3d4e31b1f195e095b699f1 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ec7e13b6f33c8978c42366b00adf9f09053e08c8 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/3/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:775e820074587e5ad868ad4dc9894b7757d8f0cb4fc8709bfdea54f392740bcb +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b93cf751a1ea95d92364b999ad2c5139cf35c7d7 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c07d7f929c4ee310bc69c3fe4ed9d3e902670d0d0743ef11de23680972f7914 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e7a9916f366c3ddabb908dc0c74d4c1fefd7947a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:309ebc629794ea6f299c08739bf14aabbc3a60f0fe4227e3099ee23ece78779a +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5e6cb487272272961ccd0be72b3dc1696a443e05 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b720bad34a829b1550cc98042b6c6b51c36f93948ab12b50482dfc3648f1d5c +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6e6b653e0c53c98a874ac32355a32660f8b27c93 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d1010d1b82e37449586ba734996c8e0212d95a29b33efe40285ce6376ba0be7 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..72d6a468684b66e9f3bd58c7c252948719e3d6f2 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:730de42a5c3d1ac7851acaaaf6c6d5f5fb9b90cedc756c36b6b46b0a54efd2f6 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9a0c5d25a38351040777d8641fe95c48cc0f7cb9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/4/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8ff4b12c44530d7ca319fc2c59c9b8163776cef48e935b3960637633252f7ba2 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d3b5f9f83c03ae9ff729809f0962786d7d15f191 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de5c49db6f4ed0cb3d01d95dd00311d061b3cb7b1c86567c7ce21183ee0f0b12 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fb31869de76306c252abcaed11ecd91b770b67fd --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dbe158f0cd865603e90dc2bd467132d18e853f72dee4f8bd29575aaa181e5a1 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8e66bfbfa9c3065a112b6dc09b1a4db230c4ce41 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5a505c65ad8691ef719edf68d2311c803ebb9825a38b6c7e1afbec976138b46 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0821c33d043d3b222bf2e9bd590a790faa3698f8 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f80b097939ebd196207775d350656ad2d4465619dc283a49dd2b5a0f5a5e92c1 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f17986e0fe448c1af6a7de2718330125f7c641f6 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:030edd3258dc2e7a12c7cf74c35fcfc5524f3d15832c403a7124fc38cc67bd9f +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a5e38de4b7fc119195b5e221f3b488083e32f1a0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/5/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d0971c136d808ad253e62d6b9f9dc2f961ee8709a12e4f087d46e41af841ec1 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..08cbacea6a47602f48b52d9454f711d0cf883953 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7802dd7c40659468bf50822897e6a3d40a144864144be7c0e430e33aacaa7fe3 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..78051e380d4c4e56d865bb11c31a0f6d7a03a56b --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:25c7d01854d2d3f72337d74776fca6829e6dc784163916c5d61bdef8799b6dd8 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d944726a0c4063f04619ffa06c748448813c5608 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:449025387a43af93068e2a4a08013e299ba12e79bda20673bb19cd32ca9c118a +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bff35030a989f2ac79c5f6414b553e45a0c3b920 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2bcd94e0a9f57ee916d30da2872be47622fd0f3e85a93a915c7b2e1a3142b175 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2764034597b3e71475ca5da931f7236a1baf4f4e --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df541469c9eb1bffb5baad8a87f887dfdf331bd44183724d159ae7e4cd56d6ef +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ff3d53412f72babb8ed00a1fef48c36e9354783f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/6/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:227cbf7fff9b926fb9086f6d06b24e52c7f62e63a22e51592dd70c20ab6504a1 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cc0b98fbe41b6fa27a3f9465757c5adec04eaede --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e0f4fb2d135fc5bf47c86624e6a6ec4addb0d8cd1740fe0aeacf60e773b9c38 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..03df284b1e8e1842287ecde430dd2c695a28ae00 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d9c5da57eccb967ae5648ab6638c65209c30296d84df16e8a2d0d1a8ae8832ec +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c63724fe6367b3b2fd6e83b5d1df34d41283713d --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60c6d1bfbf1ef5d9bd8c596690f2eb50e97263443705c5fd6a344537f56d39d7 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f270bc3e88e87838af7f1772c82019c1b8eb0e17 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ae9b4de8d562fdf5e85ba4ad3c4763dbb1520a25157c82fc5f7fe05a5f44b5a +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..37bf9bf34fc1282f598aa35312252640a3c8b253 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ecc80abf09838675c07f695444379ff839a50cdcfd7907a5cffe5ad81c1f9ffc +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a9fd1a497926f3fba5a939e33243627e5cf91fd4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/7/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e1242048bf8d447e487d6836279d77889cb1e0322857318c996bcc81219c2752 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ee5dd7bdd4dfef5b5554c59d8a3006adac0f67c0 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e254220bad9f0a123dfab657126f3d621b27df755196900a928834f358704d60 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aa91e042b520f41c770ca17b19e9731d5b15101c --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3cfa0627c671d881f47378020727054eb93f00a9306ce4ea186a51a134645d43 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7439474c4172a17ffac4806d0b5c7c278713cdc8 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:38a2414662ab70502aefcf790b9eab1efc9dac1b01aafbc3a7e96af094e46bc0 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b044f94299b383d795bfeee87a1c2489bb2092fe --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7afb2889a406d0cd4631d85cd1e0ef0e5a9c3b47f7a233af5a14b76958f37d65 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fb52201ab4766bec4fa131e470b5cb93cfc12afe --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9bdcb6f91c2d398e4a39dc65d6df7da749e58b61b9777f6bcb6322ef54aef6b5 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f1d8be33afed17187e7eca08ef263c57c01c3f20 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/8/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d8535ecb79512fbf66944ae76507d9b52d1b409dc8183f9e3aa5ae4cbde945d +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2d2b23067eb1f203897c42463e360e16d12b8649 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/attn/o_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b991551b2d4c3e79d3835e1b34ace24761b13b0f93e0980939d6fe292239ba53 +size 18874608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c67ccc363bade04a925e233488db59570a460274 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/attn/qkv_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c672467c102746680500fbc6480de2110f5714285f8bec309829aba3280f6218 +size 31457648 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/input_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/input_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..231163ceaf4914c03f5f51617dbae3aa97bc8adc --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/input_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3ae4b0c1da245821dca5c423ff5f89e995a9f79d8baadd05ad02e9e5eb4f0ea9 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..37f23a4ca696c428bd49b9be7c5dfefeda13e9ab --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/mlp/down_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae79ac24969457de15cbe604c275254761885e45131cc982fda1db82a2ac8605 +size 50331888 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..97310b7ce19a77e7263e87679ae7bfd2c8842445 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/mlp/gate_up_proj/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ca4149087f63a9836919edd9118bbc5566bab026ed5ebffc9197bb612cc2eda0 +size 100663608 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/post_attention_layernorm/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/post_attention_layernorm/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..40310a43342306a15a8ad0adf20eef7808eebc3a --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/decoder/9/pp_block/post_attention_layernorm/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2821c1384e7b7f5d3ca94750f05675f744a399c82a688ef2be437b9377420dae +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/final_layer_norm/pp_block/model_weight.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/final_layer_norm/pp_block/model_weight.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..97b253487a6c5b3d5e38641a88fec8e9c0fe1aa3 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/final_layer_norm/pp_block/model_weight.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a8b94fccc4484891767d72a3b4fb19ca230150cf5e7a1cc83a07e270a8176f56 +size 6240 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/token_position_embeddings/pp_block/token_embedding/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/token_position_embeddings/pp_block/token_embedding/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2ae502645e1c9c1a4ab2d61eb2aca369b54d60ae --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model/model/token_position_embeddings/pp_block/token_embedding/model_weight_pp-rank-0-of-2_tp-rank-0-of-1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d46824ea43abfbfb9d0f0f3fe52fe0fcb5880f849d8111b31f75fc51dc3096d +size 788005112 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model_config.json b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model_config.json new file mode 100644 index 0000000000000000000000000000000000000000..6f54ecccc80be176a7a8f7536912a7b208c9fcf5 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/model_config.json @@ -0,0 +1 @@ +{"bos_token_id": 128000, "eos_token_id": 128001, "hidden_act": "silu", "hidden_size": 3072, "initializer_range": 0.02, "intermediate_size": 8192, "is_llama_config": true, "max_position_embeddings": 4096, "num_attention_heads": 24, "num_hidden_layers": 28, "num_key_value_heads": 8, "pad_token_id": null, "pretraining_tp": 2, "rms_norm_eps": 1e-05, "rope_scaling": {"factor": 32.0, "high_freq_factor": 4.0, "low_freq_factor": 1.0, "original_max_position_embeddings": 8192, "rope_type": "llama3"}, "rope_theta": 500000.0, "rope_interleaved": false, "tie_word_embeddings": true, "use_cache": true, "vocab_size": 128256} \ No newline at end of file diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_config.json b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..264b0ec29e20f2ba890f058dbbf653ed4cd90210 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_config.json @@ -0,0 +1 @@ +{"type": "ZeroDistributedOptimizer", "parallelism": {"tp_size": "1", "dp_size": "2", "pp_size": "2", "expert_parallel_size": "1"}, "configs": {"param_name_to_dp_rank_offsets": {"model.token_position_embeddings.pp_block.token_embedding.weight": {"0": ["0", "197001216"], "1": ["197001216", "394002432"]}, "model.decoder.0.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.0.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.0.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.0.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.0.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.0.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.1.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.1.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.1.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.1.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.1.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.1.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.2.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.2.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.2.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.2.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.2.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.2.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.3.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.3.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.3.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.3.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.3.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.3.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.4.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.4.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.4.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.4.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.4.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.4.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.5.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.5.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.5.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.5.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.5.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.5.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.6.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.6.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.6.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.6.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.6.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.6.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.7.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.7.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.7.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.7.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.7.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.7.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.8.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.8.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.8.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.8.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.8.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.8.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.9.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.9.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.9.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.9.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.9.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.9.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.10.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.10.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.10.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.10.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.10.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.10.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.11.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.11.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.11.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.11.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.11.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.11.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.12.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.12.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.12.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.12.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.12.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.12.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.13.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.13.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.13.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.13.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.13.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.13.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.14.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.14.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.14.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.14.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.14.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.14.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}, "model.decoder.15.pp_block.input_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.15.pp_block.attn.qkv_proj.weight": {"0": ["0", "7864320"], "1": ["7864320", "15728640"]}, "model.decoder.15.pp_block.attn.o_proj.weight": {"0": ["0", "4718592"], "1": ["4718592", "9437184"]}, "model.decoder.15.pp_block.post_attention_layernorm.weight": {"0": ["0", "1536"], "1": ["1536", "3072"]}, "model.decoder.15.pp_block.mlp.gate_up_proj.weight": {"0": ["0", "25165824"], "1": ["25165824", "50331648"]}, "model.decoder.15.pp_block.mlp.down_proj.weight": {"0": ["0", "12582912"], "1": ["12582912", "25165824"]}}, "orig_param_shapes": {"model.token_position_embeddings.pp_block.token_embedding.weight": ["128256", "3072"], "model.decoder.0.pp_block.input_layernorm.weight": ["3072"], "model.decoder.0.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.0.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.0.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.0.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.0.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.1.pp_block.input_layernorm.weight": ["3072"], "model.decoder.1.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.1.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.1.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.1.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.1.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.2.pp_block.input_layernorm.weight": ["3072"], "model.decoder.2.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.2.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.2.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.2.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.2.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.3.pp_block.input_layernorm.weight": ["3072"], "model.decoder.3.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.3.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.3.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.3.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.3.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.4.pp_block.input_layernorm.weight": ["3072"], "model.decoder.4.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.4.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.4.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.4.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.4.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.5.pp_block.input_layernorm.weight": ["3072"], "model.decoder.5.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.5.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.5.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.5.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.5.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.6.pp_block.input_layernorm.weight": ["3072"], "model.decoder.6.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.6.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.6.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.6.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.6.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.7.pp_block.input_layernorm.weight": ["3072"], "model.decoder.7.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.7.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.7.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.7.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.7.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.8.pp_block.input_layernorm.weight": ["3072"], "model.decoder.8.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.8.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.8.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.8.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.8.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.9.pp_block.input_layernorm.weight": ["3072"], "model.decoder.9.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.9.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.9.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.9.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.9.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.10.pp_block.input_layernorm.weight": ["3072"], "model.decoder.10.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.10.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.10.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.10.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.10.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.11.pp_block.input_layernorm.weight": ["3072"], "model.decoder.11.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.11.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.11.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.11.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.11.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.12.pp_block.input_layernorm.weight": ["3072"], "model.decoder.12.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.12.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.12.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.12.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.12.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.13.pp_block.input_layernorm.weight": ["3072"], "model.decoder.13.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.13.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.13.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.13.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.13.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.14.pp_block.input_layernorm.weight": ["3072"], "model.decoder.14.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.14.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.14.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.14.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.14.pp_block.mlp.down_proj.weight": ["3072", "8192"], "model.decoder.15.pp_block.input_layernorm.weight": ["3072"], "model.decoder.15.pp_block.attn.qkv_proj.weight": ["5120", "3072"], "model.decoder.15.pp_block.attn.o_proj.weight": ["3072", "3072"], "model.decoder.15.pp_block.post_attention_layernorm.weight": ["3072"], "model.decoder.15.pp_block.mlp.gate_up_proj.weight": ["16384", "3072"], "model.decoder.15.pp_block.mlp.down_proj.weight": ["3072", "8192"]}}} \ No newline at end of file diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-0-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-0-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..a6d5a1a23e49fef78bc57e69e54e64f798df6816 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-0-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:223e4aa1192b2e4f7db5115949d684c8befc5d0b066fc6d9d04cfa83bf961c20 +size 12028411794 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-0-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-0-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..74da9336dd032841055f9b70d55ae8a99bbbb8d4 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-0-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51cc5e6322481c0d54ed648dcd68dd3f41b1dfb03a718a1f153af6e7dee4fc3f +size 12028411794 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-1-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-1-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..a6717ccf5c19716cccce1f203d05e4c1b96e2248 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-1-of-2_dp-0-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0692acb63da0c36bed9a59aadd27b4694bb1d51449b5607d176a43b793649dc8 +size 9612328561 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-1-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-1-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt new file mode 100644 index 0000000000000000000000000000000000000000..e0b6f62314caae85a4df9af569489a93a72586ec --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/optimizer/optimizer_pp-1-of-2_dp-1-of-2_tp-0-of-1_exp-0-of-1.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:41ce112f0567413344b35c777a32868e60e4e1ab2804c2ca31f3de20bf5ecd6f +size 9612328561 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-0-of-2_pp-0-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-0-of-2_pp-0-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..dc0a783730805ef41d420c8f03d8d549604c13c9 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-0-of-2_pp-0-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d81c15fba15f9a607a4736986869dc27a504e4b6e09a37603307cc1ca5aa3a52 +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-0-of-2_pp-1-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-0-of-2_pp-1-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..0de2e4a921c365e895e3b5d2820bc98c65674113 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-0-of-2_pp-1-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b19db2d5d750b0c90eae9b28695c7e22e067f62104274e6aceb33e5d8ef6ae4 +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-1-of-2_pp-0-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-1-of-2_pp-0-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c839d319130523d9bc76069b614d153897b0f2f --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-1-of-2_pp-0-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3827577ffa796c77f14fb1f3dbe7a17cff3844ab6a99a7f6240a1f2ff2b155c +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-1-of-2_pp-1-of-2.pt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-1-of-2_pp-1-of-2.pt new file mode 100644 index 0000000000000000000000000000000000000000..c4c23e4d860e23af92b26eb02a6d95d2f4784d05 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/2000/random/tp-0-of-1_dp-1-of-2_pp-1-of-2.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:900ce55cda2c6462721a7b34177d57d42e1ca1170f7f99128c90c0d1395a272f +size 1080 diff --git a/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/latest.txt b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/latest.txt new file mode 100644 index 0000000000000000000000000000000000000000..9463411b62f21b7ed88bbe711c958b3b66153330 --- /dev/null +++ b/llama-3.2-3B-finemath-4plus-p0.0-bs512-tc60B/latest.txt @@ -0,0 +1 @@ +2000 \ No newline at end of file