ChenMnZ commited on
Commit
5cfb77b
1 Parent(s): 80a0358

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # EfficientQAT
2
+
3
+ [EfficientQAT](https://arxiv.org/abs/2407.11062) is a novel quantization technical, which pushes the limitation of uniform (INT) quantization in an efficient manner. Due to the leverage of standard INT quantization, the quantized model of EfficientQAT can also be transferred into other formats, such as GPTQ, BitBLAS, etc.
4
+
5
+
6
+ In this repo, we provide three type checkpoints, one is EQAT, indicats the original checkpoints of EfficientQAT. The other two are GPTQ and BitBLAS respectively.
7
+
8
+
9
+ ## Model Zoo
10
+
11
+ We provide a number of prequantized EfficientQAT models as follows:
12
+
13
+ - WikiText2 PPL is measured in 2048 context length.
14
+ - Avg. Accuracy indicate the average accuracy in 5 zero-shot reasoning tasks (WinoGrande,PIQA,HellaSwag,Arc-Easy, Arc-Challenge) with [lm-eval v0.4.2](https://github.com/EleutherAI/lm-evaluation-harness).
15
+ - 1GB = $10^9$ Bit
16
+ - Hub Link: EQAT indicates the original checkpoints. We also transfer the checkpoints into GPTQ and BitBLAS formats, which can be loaded directly through [GPTQModel](https://github.com/ModelCloud/GPTQModel). (PS: [GPTQModel](https://github.com/ModelCloud/GPTQModel) is a official bug-fixed repo of AutoGPTQ, which would be merged into [AutoGPTQ](https://github.com/AutoGPTQ/AutoGPTQ) in future.)
17
+
18
+ | Model | Quantization | WikiText2 PPL | Avg. Accuracy | Model Size (GB) | Hub link|
19
+ |-------|--------------|---------------|---------------|-----------------|----------|
20
+ Llama-2-7B|fp16|5.47|64.86|13.2|-|
21
+ Llama-2-7B|w4g128|5.53|64.27|3.7|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-2-7b-EfficientQAT-w4g128-BitBLAS)|
22
+ Llama-2-7B|w3g128|5.81|64.02|3.1|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w3g128)|
23
+ Llama-2-7B|w2g64|6.86|60.14|2.3|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w2g64-GPTQ)\|[BitBLAS](Llama-2-7b-EfficientQAT-w2g64-BitBLAS)|
24
+ Llama-2-7B|w2g128|7.17|59.50|2.2|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-2-7b-EfficientQAT-w2g128-BitBLAS)|
25
+ Llama-2-13B|fp16|4.88|67.81|25.4|-|
26
+ Llama-2-13B|w4g128|4.93|67.52|6.8|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-13b-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-7b-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-2-7b-EfficientQAT-w4g128-BitBLAS)|
27
+ Llama-2-13B|w3g128|5.12|67.28|5.6|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-13b-EfficientQAT-w3g128)|
28
+ Llama-2-13B|w2g64|5.96|64.88|4.0|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-13b-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-13b-EfficientQAT-w2g64-GPTQ)\|[BitBLAS](Llama-2-13b-EfficientQAT-w2g64-BitBLAS)|
29
+ Llama-2-13B|w2g128|6.08|63.88|3.8|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-13b-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-13b-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-2-13b-EfficientQAT-w2g128-BitBLAS)|
30
+ Llama-2-70B|fp16|3.32|72.41|131.6|-|
31
+ Llama-2-70B|w4g128|3.39|72.62|35.8|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-2-70b-EfficientQAT-w4g128-BitBLAS)|
32
+ Llama-2-70B|w3g128|3.61|71.76|29.1|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w3g128)|
33
+ Llama-2-70B|w2g64|4.52|69.48|20.1|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w2g64-GPTQ)\|[BitBLAS](Llama-2-70b-EfficientQAT-w2g64-BitBLAS)|
34
+ Llama-2-70B|w2g128|4.61|68.93|18.9|[EQAT](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-2-70b-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-2-70b-EfficientQAT-w2g128-BitBLAS)|
35
+ Llama-3-8B|fp16|6.14|68.58|13.0|-|
36
+ Llama-3-8B|w4g128|6.47|68.43|5.4|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-3-8b-EfficientQAT-w4g128-BitBLAS)|
37
+ Llama-3-8B|w3g128|7.09|67.35|4.7|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w3g128)|
38
+ Llama-3-8B|w2g64|9.41|60.76|3.9|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-3-8b-EfficientQAT-w2g64-BitBLAS)|
39
+ Llama-3-8B|w2g128|9.80|59.36|3.8|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-8b-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-3-8b-EfficientQAT-w2g128-BitBLAS)|
40
+ Llama-3-70B|fp16|2.85|75.33|137.8|-|
41
+ Llama-3-70B|w4g128|3.17|74.57|38.9|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-3-70b-EfficientQAT-w4g128-BitBLAS)|
42
+ Llama-3-70B|w3g128|4.19|72.42|32.2|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w3g128)|
43
+ Llama-3-70B|w2g64|6.08|67.89|23.2|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w2g64-GPTQ)|
44
+ Llama-3-70B|w2g128|6.38|67.57|22.0|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-70b-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-3-70b-EfficientQAT-w2g128-BitBLAS)|
45
+ Llama-3-8B-Instruct|fp16|8.29|68.43|13.0|-|
46
+ Llama-3-8B-Instruct|w4g128|7.93|68.39|5.4|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-3-8b-instruct-EfficientQAT-w4g128-BitBLAS)|
47
+ Llama-3-8B-Instruct|w3g128|8.55|67.24|4.7|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w3g128)|
48
+ Llama-3-8B-Instruct|w2g64|11.19|60.66|3.9|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w2g64-GPTQ)\|[BitBLAS](Llama-3-8b-instruct-EfficientQAT-w2g64-BitBLAS)|
49
+ Llama-3-8B-Instruct|w2g128|11.73|60.16|3.8|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-8b-instruct-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-3-8b-instruct-EfficientQAT-w2g128-BitBLAS)|
50
+ Llama-3-70B-Instruct|fp16|5.33|73.78|137.8|-|
51
+ Llama-3-70B-Instruct|w4g128|5.35|73.47|38.9|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w4g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w4g128-GPTQ)\|[BitBLAS](Llama-3-70b-instruct-EfficientQAT-w4g128-BitBLAS)|
52
+ Llama-3-70B-Instruct|w3g128|5.65|72.87|32.2|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w3g128)|
53
+ Llama-3-70B-Instruct|w2g64|7.86|67.64|23.2|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w2g64)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w2g64-GPTQ)\|[BitBLAS](Llama-3-70b-instruct-EfficientQAT-w2g64-BitBLAS)|
54
+ Llama-3-70B-Instruct|w2g128|8.14|67.54|22.0|[EQAT](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w2g128)\|[GPTQ](https://huggingface.co/ChenMnZ/Llama-3-70b-instruct-EfficientQAT-w2g128-GPTQ)\|[BitBLAS](Llama-3-70b-instruct-EfficientQAT-w2g128-BitBLAS)|
55
+
56
+ ## Usage of EQAT models
57
+ Please refer [https://github.com/OpenGVLab/EfficientQAT](https://github.com/OpenGVLab/EfficientQAT?tab=readme-ov-file#inference) for details.
58
+
59
+ ## Usage of GPTQ and BitBLAS models
60
+ Below is an example to inference with GPTQ or BitBLAS quantized formats.
61
+ ```Python
62
+ from transformers import AutoTokenizer
63
+ from gptqmodel import GPTQModel
64
+
65
+ quant_dir = "ChenMnZ/Llama-2-7b-EfficientQAT-w2g128-GPTQ"
66
+ # quant_dir = "ChenMnZ/Llama-2-7b-EfficientQAT-w2g128-BitBLAS"
67
+ # or local path
68
+
69
+ tokenizer = AutoTokenizer.from_pretrained(quant_dir, use_fast=True)
70
+
71
+
72
+ # load quantized model to the first GPU
73
+ model = GPTQModel.from_quantized(quant_dir)
74
+
75
+ # inference with model.generate
76
+ print(tokenizer.decode(model.generate(**tokenizer("Model quantization is", return_tensors="pt").to(model.device))[0]))
77
+ ```
78
+
79
+
80
+ ## Citation
81
+ If you found this work useful, please consider citing:
82
+ ```
83
+ @article{efficientqat,
84
+ title={EfficientQAT: Efficient Quantization-Aware Training for Large Language Models},
85
+ author={Chen, Mengzhao and Shao, Wenqi and Xu, Peng and Wang, Jiahao and Gao, Peng and Zhang, Kaipeng and Qiao, Yu and Luo, Ping},
86
+ journal={arXiv preprint arXiv:2407.11062},
87
+ year={2024}
88
+ }
89
+ ```
config.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/cpfs01/user/chenmengzhao/efficientqat_repo/efficientqat_checkpoints_original/Llama-2-7b-EfficientQAT-w2g64",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 4096,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 11008,
14
+ "max_position_embeddings": 4096,
15
+ "mlp_bias": false,
16
+ "model_type": "llama",
17
+ "num_attention_heads": 32,
18
+ "num_hidden_layers": 32,
19
+ "num_key_value_heads": 32,
20
+ "pad_token_id": 0,
21
+ "pretraining_tp": 1,
22
+ "quantization_config": {
23
+ "bits": 2,
24
+ "checkpoint_format": "gptq_v2",
25
+ "damp_percent": 0.01,
26
+ "desc_act": false,
27
+ "group_size": 64,
28
+ "lm_head": false,
29
+ "meta": {
30
+ "quantizer": "gptqmodel:0.9.9-dev0"
31
+ },
32
+ "model_file_base_name": "model",
33
+ "model_name_or_path": "/cpfs01/user/chenmengzhao/efficientqat_repo/efficientqat_checkpoints_original/Llama-2-7b-EfficientQAT-w2g64",
34
+ "quant_method": "gptq",
35
+ "static_groups": false,
36
+ "sym": false,
37
+ "true_sequential": true
38
+ },
39
+ "rms_norm_eps": 1e-05,
40
+ "rope_scaling": null,
41
+ "rope_theta": 10000.0,
42
+ "tie_word_embeddings": false,
43
+ "torch_dtype": "float16",
44
+ "transformers_version": "4.42.4",
45
+ "use_cache": false,
46
+ "vocab_size": 32000
47
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:88ffd82aa9f41e51e91b119b15324a9120e5a7334b95f59d20c798e1a4ed61d0
3
+ size 2376156344
quantize_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 2,
3
+ "group_size": 64,
4
+ "desc_act": false,
5
+ "static_groups": false,
6
+ "sym": false,
7
+ "lm_head": false,
8
+ "damp_percent": 0.01,
9
+ "true_sequential": true,
10
+ "model_name_or_path": "/cpfs01/user/chenmengzhao/efficientqat_repo/efficientqat_checkpoints_GPTQ/Llama-2-7b-EfficientQAT-w2g64-GPTQ",
11
+ "model_file_base_name": "model",
12
+ "quant_method": "gptq",
13
+ "checkpoint_format": "gptq_v2",
14
+ "meta": {
15
+ "quantizer": "gptqmodel:0.9.9-dev0"
16
+ }
17
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "unk_token": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
tokenizer_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": true,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<unk>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<s>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "2": {
23
+ "content": "</s>",
24
+ "lstrip": false,
25
+ "normalized": false,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": true
29
+ }
30
+ },
31
+ "bos_token": "<s>",
32
+ "clean_up_tokenization_spaces": false,
33
+ "eos_token": "</s>",
34
+ "legacy": false,
35
+ "model_max_length": 4096,
36
+ "pad_token": null,
37
+ "padding_side": "right",
38
+ "sp_model_kwargs": {},
39
+ "spaces_between_special_tokens": false,
40
+ "tokenizer_class": "LlamaTokenizer",
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": false
43
+ }