decoder full weights: 60M dense-Qwen3.5 GDN next-word predictor (HF format, fp32 safetensors + tokenizer)
Browse files
pred_q35_60m/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3_5ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"head_dim": 64,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 640,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 2560,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"linear_attention",
|
| 17 |
+
"linear_attention",
|
| 18 |
+
"linear_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"linear_attention",
|
| 21 |
+
"linear_attention",
|
| 22 |
+
"linear_attention",
|
| 23 |
+
"full_attention"
|
| 24 |
+
],
|
| 25 |
+
"linear_conv_kernel_dim": 4,
|
| 26 |
+
"linear_key_head_dim": 128,
|
| 27 |
+
"linear_num_key_heads": 2,
|
| 28 |
+
"linear_num_value_heads": 4,
|
| 29 |
+
"linear_value_head_dim": 128,
|
| 30 |
+
"max_position_embeddings": 2048,
|
| 31 |
+
"model_type": "qwen3_5_text",
|
| 32 |
+
"num_attention_heads": 8,
|
| 33 |
+
"num_hidden_layers": 8,
|
| 34 |
+
"num_key_value_heads": 2,
|
| 35 |
+
"pad_token_id": null,
|
| 36 |
+
"partial_rotary_factor": 0.25,
|
| 37 |
+
"rms_norm_eps": 1e-06,
|
| 38 |
+
"rope_parameters": {
|
| 39 |
+
"partial_rotary_factor": 0.25,
|
| 40 |
+
"rope_theta": 10000.0,
|
| 41 |
+
"rope_type": "default"
|
| 42 |
+
},
|
| 43 |
+
"tie_word_embeddings": true,
|
| 44 |
+
"transformers_version": "5.13.1",
|
| 45 |
+
"use_cache": true,
|
| 46 |
+
"vocab_size": 16000
|
| 47 |
+
}
|
pred_q35_60m/generation_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"output_attentions": false,
|
| 6 |
+
"output_hidden_states": false,
|
| 7 |
+
"transformers_version": "5.13.1",
|
| 8 |
+
"use_cache": true
|
| 9 |
+
}
|
pred_q35_60m/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6700f0a19fefb6e7cfecded08bd6ba153f4d601c573a810cb048aa0e7f0e648
|
| 3 |
+
size 119585784
|
pred_q35_60m/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|