Feature Extraction
Transformers
Safetensors
qwen3
speculative-decoding
diffusion-drafter
adaflash
custom_code
Instructions to use AdaFlash/Qwen3-8B-AdaFlash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use AdaFlash/Qwen3-8B-AdaFlash with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="AdaFlash/Qwen3-8B-AdaFlash", trust_remote_code=True)# Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("AdaFlash/Qwen3-8B-AdaFlash", trust_remote_code=True) model = AutoModel.from_pretrained("AdaFlash/Qwen3-8B-AdaFlash", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Upload folder using huggingface_hub
Browse files- config.json +55 -0
- model.safetensors +3 -0
- training_state.pt +3 -0
- version.txt +1 -0
config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"DFlashDraftModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoModel": "dflash.DFlashDraftModel"
|
| 9 |
+
},
|
| 10 |
+
"block_size": 16,
|
| 11 |
+
"bos_token_id": 151643,
|
| 12 |
+
"dflash_config": {
|
| 13 |
+
"mask_token_id": 151669,
|
| 14 |
+
"target_layer_ids": [
|
| 15 |
+
1,
|
| 16 |
+
9,
|
| 17 |
+
17,
|
| 18 |
+
25,
|
| 19 |
+
33
|
| 20 |
+
],
|
| 21 |
+
"thresh_head_bottleneck_dim": 256,
|
| 22 |
+
"thresh_head_direct_len": true,
|
| 23 |
+
"use_thresh_head_two_model": true
|
| 24 |
+
},
|
| 25 |
+
"dtype": "bfloat16",
|
| 26 |
+
"eos_token_id": 151645,
|
| 27 |
+
"head_dim": 128,
|
| 28 |
+
"hidden_act": "silu",
|
| 29 |
+
"hidden_size": 4096,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 12288,
|
| 32 |
+
"layer_types": [
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention"
|
| 38 |
+
],
|
| 39 |
+
"max_position_embeddings": 40960,
|
| 40 |
+
"max_window_layers": 5,
|
| 41 |
+
"model_type": "qwen3",
|
| 42 |
+
"num_attention_heads": 32,
|
| 43 |
+
"num_hidden_layers": 5,
|
| 44 |
+
"num_key_value_heads": 8,
|
| 45 |
+
"num_target_layers": 36,
|
| 46 |
+
"rms_norm_eps": 1e-06,
|
| 47 |
+
"rope_scaling": null,
|
| 48 |
+
"rope_theta": 1000000,
|
| 49 |
+
"sliding_window": null,
|
| 50 |
+
"tie_word_embeddings": false,
|
| 51 |
+
"transformers_version": "4.57.1",
|
| 52 |
+
"use_cache": true,
|
| 53 |
+
"use_sliding_window": false,
|
| 54 |
+
"vocab_size": 151936
|
| 55 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc73adba77efb79f3a22bc23d734f894d8d86dd2bb4a68a44aeeed856c57e74c
|
| 3 |
+
size 2101586658
|
training_state.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea3e06f19f4a8ac5f0af3316c17ac948a58b77f32f978741c9711e91c13f1b01
|
| 3 |
+
size 1309002021
|
version.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
1
|