hik63382 commited on
Commit
afa5352
·
verified ·
1 Parent(s): 9e89d98

Upload 8 files

Browse files
.gitattributes CHANGED
@@ -41,3 +41,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
41
  002_Eng-turbo/onnx/embed_tokens_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
42
  002_Eng-turbo/onnx/language_model_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
43
  002_Eng-turbo/onnx/speech_encoder_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
 
 
41
  002_Eng-turbo/onnx/embed_tokens_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
42
  002_Eng-turbo/onnx/language_model_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
43
  002_Eng-turbo/onnx/speech_encoder_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
44
+ 002_Eng-turbo/high-1764639553.jpg filter=lfs diff=lfs merge=lfs -text
002_Eng-turbo/README.md ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ base_model: ResembleAI/chatterbox-turbo-ONNX
4
+ tags:
5
+ - text-to-speech
6
+ - tts
7
+ - onnx
8
+ - webgpu
9
+ - transformers.js
10
+ ---
11
+
12
+ # Chatterbox Turbo - WebGPU Compatible
13
+
14
+ This is a WebGPU-compatible version of [ResembleAI/chatterbox-turbo-ONNX](https://huggingface.co/ResembleAI/chatterbox-turbo-ONNX).
15
+
16
+ ## Changes from Original
17
+
18
+ The original model contains `int64` Cast operations and tensors that WebGPU cannot execute.
19
+ This version converts all `int64` operations to `int32`, enabling direct WebGPU inference.
20
+
21
+ ### Modifications Made:
22
+ - **conditional_decoder**: 521 Cast nodes inserted (376 Shape/Range ops)
23
+ - **speech_encoder**: 350 Cast nodes inserted (243 Shape/Range ops)
24
+ - **language_model**: 3 Cast nodes inserted
25
+ - **embed_tokens**: 1 Cast node inserted
26
+
27
+ ## Usage with Transformers.js
28
+
29
+ ```javascript
30
+ import { AutoModel, AutoProcessor } from '@huggingface/transformers';
31
+
32
+ const model = await AutoModel.from_pretrained('spacekaren/chatterbox-turbo-webgpu', {
33
+ device: 'webgpu',
34
+ dtype: 'q4f16',
35
+ });
36
+
37
+ const processor = await AutoProcessor.from_pretrained('spacekaren/chatterbox-turbo-webgpu');
38
+ ```
39
+
40
+ ## Model Size
41
+
42
+ - **Total**: ~539 MB (q4f16 quantization)
43
+ - Same architecture as original, just int64→int32 conversion
44
+
45
+ ## License
46
+
47
+ MIT (same as original)
48
+
49
+ ## Credits
50
+
51
+ - Original model: [ResembleAI/chatterbox-turbo-ONNX](https://huggingface.co/ResembleAI/chatterbox-turbo-ONNX)
52
+ - Conversion script: [local.core/scripts/convert_int64_to_int32.py](https://github.com/anthropics/lama)
002_Eng-turbo/config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "chatterbox",
3
+ "text_config": {
4
+ "activation_function": "gelu_new",
5
+ "architectures": [
6
+ "GPT2LMHeadModel"
7
+ ],
8
+ "attn_pdrop": 0.1,
9
+ "bos_token_id": 50256,
10
+ "embd_pdrop": 0.1,
11
+ "eos_token_id": 50256,
12
+ "initializer_range": 0.02,
13
+ "layer_norm_epsilon": 1e-05,
14
+ "model_type": "gpt2",
15
+ "n_ctx": 8196,
16
+ "n_embd": 1024,
17
+ "n_head": 16,
18
+ "n_inner": null,
19
+ "n_layer": 24,
20
+ "n_positions": 8196,
21
+ "n_special": 0,
22
+ "predict_special_tokens": true,
23
+ "reorder_and_upcast_attn": false,
24
+ "resid_pdrop": 0.1,
25
+ "scale_attn_by_inverse_layer_idx": false,
26
+ "scale_attn_weights": true,
27
+ "summary_activation": null,
28
+ "summary_first_dropout": 0.1,
29
+ "summary_proj_to_labels": true,
30
+ "summary_type": "cls_index",
31
+ "summary_use_proj": true,
32
+ "task_specific_params": {
33
+ "text-generation": {
34
+ "do_sample": true,
35
+ "max_length": 50
36
+ }
37
+ },
38
+ "transformers_version": "5.0.0.dev0",
39
+ "use_cache": true,
40
+ "vocab_size": 6563
41
+ },
42
+ "transformers.js_config": {
43
+ "use_external_data_format": true,
44
+ "kv_cache_dtype": {
45
+ "fp16": "float16",
46
+ "q4f16": "float16"
47
+ }
48
+ }
49
+ }
002_Eng-turbo/generation_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "eos_token_id": 6562,
3
+ "repetition_penalty": 1.2
4
+ }
002_Eng-turbo/high-1764639553.jpg ADDED

Git LFS Details

  • SHA256: 94c318266c496e74e2f1ce0d1187c4ebc6714ba15fb94ce10811880c3ad4ea37
  • Pointer size: 131 Bytes
  • Size of remote file: 129 kB
002_Eng-turbo/nzg nzg 73.txt ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+
3
+
4
+ 📺 YouTube :- @NZG73
5
+
6
+
7
+ 🌐 Website:- nzg73.blogspot.com
8
+
9
+
10
+ 📧 Email:- nzgnzg73@gmail.com
11
+
12
+
002_Eng-turbo/preprocessor_config.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "feature_extractor_type": "ChatterboxFeatureExtractor",
3
+ "processor_class": "ChatterboxProcessor",
4
+ "sampling_rate": 24000
5
+ }
002_Eng-turbo/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
002_Eng-turbo/tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "additional_special_tokens": null,
4
+ "backend": "tokenizers",
5
+ "bos_token": "<|endoftext|>",
6
+ "clean_up_tokenization_spaces": false,
7
+ "eos_token": "<|endoftext|>",
8
+ "errors": "replace",
9
+ "is_local": false,
10
+ "model_max_length": 1024,
11
+ "model_specific_special_tokens": {},
12
+ "pad_token": "<|endoftext|>",
13
+ "tokenizer_class": "GPT2Tokenizer",
14
+ "unk_token": "<|endoftext|>"
15
+ }