fbaldassarri
commited on
Commit
•
6a56195
1
Parent(s):
eb589cd
Initial Upload
Browse files- README.md +93 -3
- config.json +54 -0
- generation_config.json +6 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- quantization_config.json +24 -0
- special_tokens_map.json +42 -0
- tokenizer.json +0 -0
- tokenizer_config.json +168 -0
- vocab.json +0 -0
README.md
CHANGED
@@ -1,3 +1,93 @@
|
|
1 |
-
---
|
2 |
-
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- en
|
4 |
+
- de
|
5 |
+
- fr
|
6 |
+
- it
|
7 |
+
- pt
|
8 |
+
- hi
|
9 |
+
- es
|
10 |
+
- th
|
11 |
+
license: apache-2.0
|
12 |
+
library_name: transformers
|
13 |
+
tags:
|
14 |
+
- autoround
|
15 |
+
- auto-round
|
16 |
+
- intel-autoround
|
17 |
+
- gptq
|
18 |
+
- woq
|
19 |
+
- pytorch
|
20 |
+
- transformers
|
21 |
+
- intel
|
22 |
+
model_name: SmolLM2 1.7B
|
23 |
+
base_model: HuggingFaceTB/SmolLM2-1.7B
|
24 |
+
inference: false
|
25 |
+
model_creator: HuggingFaceTB
|
26 |
+
pipeline_tag: text-generation
|
27 |
+
prompt_template: '{prompt}
|
28 |
+
'
|
29 |
+
quantized_by: fbaldassarri
|
30 |
+
---
|
31 |
+
|
32 |
+
## Model Information
|
33 |
+
|
34 |
+
Quantized version of [HuggingFaceTB/SmolLM2-1.7B](HuggingFaceTB/SmolLM2-1.7B) using torch.float32 for quantization tuning.
|
35 |
+
- 4 bits (INT4)
|
36 |
+
- group size = 128
|
37 |
+
- Symmetrical Quantization
|
38 |
+
- Method WoQ (AutoRound format)
|
39 |
+
|
40 |
+
Fast and low memory, 2-3X speedup (slight accuracy drop at W4G128)
|
41 |
+
|
42 |
+
Quantization framework: [Intel AutoRound](https://github.com/intel/auto-round) v0.4.2
|
43 |
+
|
44 |
+
Note: this INT4 version of SmolLM2-1.7B has been quantized to run inference through CPU.
|
45 |
+
|
46 |
+
## Replication Recipe
|
47 |
+
|
48 |
+
### Step 1 Install Requirements
|
49 |
+
|
50 |
+
I suggest to install requirements into a dedicated python-virtualenv or a conda enviroment.
|
51 |
+
|
52 |
+
```
|
53 |
+
python -m pip install <package> --upgrade
|
54 |
+
```
|
55 |
+
|
56 |
+
- accelerate==1.2.0
|
57 |
+
- autoawq==0.2.7.post3
|
58 |
+
- auto_gptq==0.7.1
|
59 |
+
- neural_compressor==3.1.1
|
60 |
+
- torch==2.4.1+cpu
|
61 |
+
- torchaudio==2.4.1+cpu
|
62 |
+
- torchvision==0.19.1+cpu
|
63 |
+
- transformers==4.47.0
|
64 |
+
|
65 |
+
### Step 2 Build Intel Autoround wheel from sources
|
66 |
+
|
67 |
+
```
|
68 |
+
python -m pip install git+https://github.com/intel/auto-round.git
|
69 |
+
```
|
70 |
+
|
71 |
+
### Step 3 Script for Quantization
|
72 |
+
|
73 |
+
```
|
74 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
75 |
+
model_name = "HuggingFaceTB/SmolLM2-1.7B"
|
76 |
+
model = AutoModelForCausalLM.from_pretrained(model_name)
|
77 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
78 |
+
from auto_round import AutoRound
|
79 |
+
bits, group_size, sym, device, amp = 4, 128, True, 'cpu', False
|
80 |
+
autoround = AutoRound(model, tokenizer, nsamples=128, iters=200, seqlen=512, batch_size=4, bits=bits, group_size=group_size, sym=sym, device=device, amp=amp)
|
81 |
+
autoround.quantize()
|
82 |
+
output_dir = "./AutoRound/HuggingFaceTB_SmolLM2-1.7B-auto_round-int4-gs128-sym"
|
83 |
+
autoround.save_quantized(output_dir, format='auto_round', inplace=True)
|
84 |
+
```
|
85 |
+
|
86 |
+
## License
|
87 |
+
|
88 |
+
[Apache 2.0 License](https://choosealicense.com/licenses/apache-2.0/)
|
89 |
+
|
90 |
+
## Disclaimer
|
91 |
+
|
92 |
+
This quantized model comes with no warrenty. It has been developed only for research purposes.
|
93 |
+
|
config.json
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "HuggingFaceTB/SmolLM2-1.7B",
|
3 |
+
"architectures": [
|
4 |
+
"LlamaForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"bos_token_id": 0,
|
9 |
+
"eos_token_id": 0,
|
10 |
+
"head_dim": 64,
|
11 |
+
"hidden_act": "silu",
|
12 |
+
"hidden_size": 2048,
|
13 |
+
"initializer_range": 0.02,
|
14 |
+
"intermediate_size": 8192,
|
15 |
+
"max_position_embeddings": 8192,
|
16 |
+
"mlp_bias": false,
|
17 |
+
"model_type": "llama",
|
18 |
+
"num_attention_heads": 32,
|
19 |
+
"num_hidden_layers": 24,
|
20 |
+
"num_key_value_heads": 32,
|
21 |
+
"pretraining_tp": 1,
|
22 |
+
"quantization_config": {
|
23 |
+
"amp": false,
|
24 |
+
"autoround_version": "0.4.2",
|
25 |
+
"backend": "auto_round:gptq:exllamav2",
|
26 |
+
"batch_size": 4,
|
27 |
+
"bits": 4,
|
28 |
+
"data_type": "int",
|
29 |
+
"dataset": "NeelNanda/pile-10k",
|
30 |
+
"enable_minmax_tuning": true,
|
31 |
+
"enable_norm_bias_tuning": false,
|
32 |
+
"enable_quanted_input": true,
|
33 |
+
"gradient_accumulate_steps": 1,
|
34 |
+
"group_size": 128,
|
35 |
+
"iters": 200,
|
36 |
+
"low_gpu_mem_usage": false,
|
37 |
+
"lr": 0.005,
|
38 |
+
"minmax_lr": 0.005,
|
39 |
+
"nsamples": 128,
|
40 |
+
"quant_method": "intel/auto-round",
|
41 |
+
"scale_dtype": "torch.float16",
|
42 |
+
"seqlen": 512,
|
43 |
+
"sym": true,
|
44 |
+
"to_quant_block_names": null
|
45 |
+
},
|
46 |
+
"rms_norm_eps": 1e-05,
|
47 |
+
"rope_scaling": null,
|
48 |
+
"rope_theta": 130000,
|
49 |
+
"tie_word_embeddings": true,
|
50 |
+
"torch_dtype": "float32",
|
51 |
+
"transformers_version": "4.47.0",
|
52 |
+
"use_cache": true,
|
53 |
+
"vocab_size": 49152
|
54 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 0,
|
4 |
+
"eos_token_id": 0,
|
5 |
+
"transformers_version": "4.47.0"
|
6 |
+
}
|
merges.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b1d861bf749a2b114941bb8d477ebcee046a67651fd826ca37ed70826ecf5c97
|
3 |
+
size 1241864760
|
quantization_config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bits": 4,
|
3 |
+
"group_size": 128,
|
4 |
+
"sym": true,
|
5 |
+
"data_type": "int",
|
6 |
+
"enable_quanted_input": true,
|
7 |
+
"enable_minmax_tuning": true,
|
8 |
+
"seqlen": 512,
|
9 |
+
"batch_size": 4,
|
10 |
+
"scale_dtype": "torch.float16",
|
11 |
+
"lr": 0.005,
|
12 |
+
"minmax_lr": 0.005,
|
13 |
+
"gradient_accumulate_steps": 1,
|
14 |
+
"iters": 200,
|
15 |
+
"amp": false,
|
16 |
+
"nsamples": 128,
|
17 |
+
"low_gpu_mem_usage": false,
|
18 |
+
"to_quant_block_names": null,
|
19 |
+
"enable_norm_bias_tuning": false,
|
20 |
+
"dataset": "NeelNanda/pile-10k",
|
21 |
+
"autoround_version": "0.4.2",
|
22 |
+
"quant_method": "intel/auto-round",
|
23 |
+
"backend": "auto_round:gptq:exllamav2"
|
24 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"<|endoftext|>",
|
4 |
+
"<|im_start|>",
|
5 |
+
"<|im_end|>",
|
6 |
+
"<repo_name>",
|
7 |
+
"<reponame>",
|
8 |
+
"<file_sep>",
|
9 |
+
"<filename>",
|
10 |
+
"<gh_stars>",
|
11 |
+
"<issue_start>",
|
12 |
+
"<issue_comment>",
|
13 |
+
"<issue_closed>",
|
14 |
+
"<jupyter_start>",
|
15 |
+
"<jupyter_text>",
|
16 |
+
"<jupyter_code>",
|
17 |
+
"<jupyter_output>",
|
18 |
+
"<jupyter_script>",
|
19 |
+
"<empty_output>"
|
20 |
+
],
|
21 |
+
"bos_token": {
|
22 |
+
"content": "<|endoftext|>",
|
23 |
+
"lstrip": false,
|
24 |
+
"normalized": false,
|
25 |
+
"rstrip": false,
|
26 |
+
"single_word": false
|
27 |
+
},
|
28 |
+
"eos_token": {
|
29 |
+
"content": "<|endoftext|>",
|
30 |
+
"lstrip": false,
|
31 |
+
"normalized": false,
|
32 |
+
"rstrip": false,
|
33 |
+
"single_word": false
|
34 |
+
},
|
35 |
+
"unk_token": {
|
36 |
+
"content": "<|endoftext|>",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false
|
41 |
+
}
|
42 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,168 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_prefix_space": false,
|
3 |
+
"added_tokens_decoder": {
|
4 |
+
"0": {
|
5 |
+
"content": "<|endoftext|>",
|
6 |
+
"lstrip": false,
|
7 |
+
"normalized": false,
|
8 |
+
"rstrip": false,
|
9 |
+
"single_word": false,
|
10 |
+
"special": true
|
11 |
+
},
|
12 |
+
"1": {
|
13 |
+
"content": "<|im_start|>",
|
14 |
+
"lstrip": false,
|
15 |
+
"normalized": false,
|
16 |
+
"rstrip": false,
|
17 |
+
"single_word": false,
|
18 |
+
"special": true
|
19 |
+
},
|
20 |
+
"2": {
|
21 |
+
"content": "<|im_end|>",
|
22 |
+
"lstrip": false,
|
23 |
+
"normalized": false,
|
24 |
+
"rstrip": false,
|
25 |
+
"single_word": false,
|
26 |
+
"special": true
|
27 |
+
},
|
28 |
+
"3": {
|
29 |
+
"content": "<repo_name>",
|
30 |
+
"lstrip": false,
|
31 |
+
"normalized": false,
|
32 |
+
"rstrip": false,
|
33 |
+
"single_word": false,
|
34 |
+
"special": true
|
35 |
+
},
|
36 |
+
"4": {
|
37 |
+
"content": "<reponame>",
|
38 |
+
"lstrip": false,
|
39 |
+
"normalized": false,
|
40 |
+
"rstrip": false,
|
41 |
+
"single_word": false,
|
42 |
+
"special": true
|
43 |
+
},
|
44 |
+
"5": {
|
45 |
+
"content": "<file_sep>",
|
46 |
+
"lstrip": false,
|
47 |
+
"normalized": false,
|
48 |
+
"rstrip": false,
|
49 |
+
"single_word": false,
|
50 |
+
"special": true
|
51 |
+
},
|
52 |
+
"6": {
|
53 |
+
"content": "<filename>",
|
54 |
+
"lstrip": false,
|
55 |
+
"normalized": false,
|
56 |
+
"rstrip": false,
|
57 |
+
"single_word": false,
|
58 |
+
"special": true
|
59 |
+
},
|
60 |
+
"7": {
|
61 |
+
"content": "<gh_stars>",
|
62 |
+
"lstrip": false,
|
63 |
+
"normalized": false,
|
64 |
+
"rstrip": false,
|
65 |
+
"single_word": false,
|
66 |
+
"special": true
|
67 |
+
},
|
68 |
+
"8": {
|
69 |
+
"content": "<issue_start>",
|
70 |
+
"lstrip": false,
|
71 |
+
"normalized": false,
|
72 |
+
"rstrip": false,
|
73 |
+
"single_word": false,
|
74 |
+
"special": true
|
75 |
+
},
|
76 |
+
"9": {
|
77 |
+
"content": "<issue_comment>",
|
78 |
+
"lstrip": false,
|
79 |
+
"normalized": false,
|
80 |
+
"rstrip": false,
|
81 |
+
"single_word": false,
|
82 |
+
"special": true
|
83 |
+
},
|
84 |
+
"10": {
|
85 |
+
"content": "<issue_closed>",
|
86 |
+
"lstrip": false,
|
87 |
+
"normalized": false,
|
88 |
+
"rstrip": false,
|
89 |
+
"single_word": false,
|
90 |
+
"special": true
|
91 |
+
},
|
92 |
+
"11": {
|
93 |
+
"content": "<jupyter_start>",
|
94 |
+
"lstrip": false,
|
95 |
+
"normalized": false,
|
96 |
+
"rstrip": false,
|
97 |
+
"single_word": false,
|
98 |
+
"special": true
|
99 |
+
},
|
100 |
+
"12": {
|
101 |
+
"content": "<jupyter_text>",
|
102 |
+
"lstrip": false,
|
103 |
+
"normalized": false,
|
104 |
+
"rstrip": false,
|
105 |
+
"single_word": false,
|
106 |
+
"special": true
|
107 |
+
},
|
108 |
+
"13": {
|
109 |
+
"content": "<jupyter_code>",
|
110 |
+
"lstrip": false,
|
111 |
+
"normalized": false,
|
112 |
+
"rstrip": false,
|
113 |
+
"single_word": false,
|
114 |
+
"special": true
|
115 |
+
},
|
116 |
+
"14": {
|
117 |
+
"content": "<jupyter_output>",
|
118 |
+
"lstrip": false,
|
119 |
+
"normalized": false,
|
120 |
+
"rstrip": false,
|
121 |
+
"single_word": false,
|
122 |
+
"special": true
|
123 |
+
},
|
124 |
+
"15": {
|
125 |
+
"content": "<jupyter_script>",
|
126 |
+
"lstrip": false,
|
127 |
+
"normalized": false,
|
128 |
+
"rstrip": false,
|
129 |
+
"single_word": false,
|
130 |
+
"special": true
|
131 |
+
},
|
132 |
+
"16": {
|
133 |
+
"content": "<empty_output>",
|
134 |
+
"lstrip": false,
|
135 |
+
"normalized": false,
|
136 |
+
"rstrip": false,
|
137 |
+
"single_word": false,
|
138 |
+
"special": true
|
139 |
+
}
|
140 |
+
},
|
141 |
+
"additional_special_tokens": [
|
142 |
+
"<|endoftext|>",
|
143 |
+
"<|im_start|>",
|
144 |
+
"<|im_end|>",
|
145 |
+
"<repo_name>",
|
146 |
+
"<reponame>",
|
147 |
+
"<file_sep>",
|
148 |
+
"<filename>",
|
149 |
+
"<gh_stars>",
|
150 |
+
"<issue_start>",
|
151 |
+
"<issue_comment>",
|
152 |
+
"<issue_closed>",
|
153 |
+
"<jupyter_start>",
|
154 |
+
"<jupyter_text>",
|
155 |
+
"<jupyter_code>",
|
156 |
+
"<jupyter_output>",
|
157 |
+
"<jupyter_script>",
|
158 |
+
"<empty_output>"
|
159 |
+
],
|
160 |
+
"bos_token": "<|endoftext|>",
|
161 |
+
"clean_up_tokenization_spaces": false,
|
162 |
+
"eos_token": "<|endoftext|>",
|
163 |
+
"extra_special_tokens": {},
|
164 |
+
"model_max_length": 8192,
|
165 |
+
"tokenizer_class": "GPT2Tokenizer",
|
166 |
+
"unk_token": "<|endoftext|>",
|
167 |
+
"vocab_size": 49152
|
168 |
+
}
|
vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|