kuotient commited on
Commit
1aaab80
โ€ข
1 Parent(s): 2389f6a

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +34 -13
README.md CHANGED
@@ -19,25 +19,45 @@ tags:
19
  - gptq
20
  license: cc-by-nc-sa-4.0
21
  ---
22
- # WIP
23
- ## Llama-2-Ko-GPTQ
 
 
 
 
 
 
 
 
 
 
24
  <!-- README_GPTQ.md-provided-files start -->
25
  ## Provided files and GPTQ parameters
26
- Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements.
27
- Each separate quant is in a different branch. See below for instructions on fetching from different branches.
28
- All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches are made with AutoGPTQ. Files in the `main` branch which were uploaded before August 2023 were made with GPTQ-for-LLaMa.
 
29
  <details>
30
- <summary>Explanation of GPTQ parameters</summary>
31
- - Bits: The bit size of the quantised model.
32
- - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value.
33
- - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now.
34
- - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy.
35
- - GPTQ dataset: The dataset used for quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s).
36
- - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences.
37
- - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama models in 4-bit.
 
38
 
39
  </details>
40
 
 
 
 
 
 
 
 
 
41
  # Original model card: Llama 2 ko 70b
42
  > ๐Ÿšง Note: this repo is under construction ๐Ÿšง
43
 
@@ -167,3 +187,4 @@ Apple Sillicon does not support BF16 computing, use CPU instead. (BF16 is suppor
167
  publisher = { Hugging Face }
168
  }
169
  ```
 
 
19
  - gptq
20
  license: cc-by-nc-sa-4.0
21
  ---
22
+
23
+ # Llama-2-Ko-70b-GPTQ
24
+ - ๋ชจ๋ธ ์ œ์ž‘์ž: [beomi](https://huggingface.co/beomi)
25
+ - ์›๋ณธ ๋ชจ๋ธ: [Llama-2-ko-70b](https://huggingface.co/beomi/llama-2-ko-70b)
26
+
27
+ <!-- description start -->
28
+ ## Description
29
+
30
+ ์ด ๋ ˆํฌ๋Š” [Llama-2-ko-70b](https://huggingface.co/beomi/llama-2-ko-70b)์˜ GPTQ ๋ชจ๋ธ ํŒŒ์ผ์„ ํฌํ•จํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
31
+
32
+ <!-- description end -->
33
+
34
  <!-- README_GPTQ.md-provided-files start -->
35
  ## Provided files and GPTQ parameters
36
+ ํ•˜๋“œ์›จ์–ด์™€ ์š”๊ตฌ์‚ฌํ•ญ์— ๊ฐ€์žฅ ์ ํ•ฉํ•œ ์–‘์žํ™” ๋งค๊ฐœ๋ณ€์ˆ˜๋ฅผ ์„ ํƒํ•  ์ˆ˜ ์žˆ๋„๋ก ์—ฌ๋Ÿฌ ๊ฐ€์ง€(๊ณง) ์–‘์žํ™” ๋งค๊ฐœ๋ณ€์ˆ˜๊ฐ€ ์ œ๊ณต๋ฉ๋‹ˆ๋‹ค.
37
+ ๊ฐ ์–‘์žํ™”๋Š” ๋‹ค๋ฅธ ๋ธŒ๋žœ์น˜์— ์žˆ์Šต๋‹ˆ๋‹ค.
38
+ ๋ชจ๋“  GPTQ ์–‘์žํ™”๋Š” AutoGPTQ๋กœ ๋งŒ๋“ค์–ด์กŒ์Šต๋‹ˆ๋‹ค.
39
+
40
  <details>
41
+ <summary>GPTQ ํŒŒ๋ผ๋ฏธํ„ฐ ์ •๋ณด</summary>
42
+
43
+ - Bits: ์–‘์žํ™”๋œ ๋ชจ๋ธ์˜ ๋น„ํŠธ ํฌ๊ธฐ์ž…๋‹ˆ๋‹ค.
44
+ - GS: GPTQ ๊ทธ๋ฃน ์‚ฌ์ด์ฆˆ. ์ˆซ์ž๊ฐ€ ๋†’์„์ˆ˜๋ก VRAM์„ ๋œ ์‚ฌ์šฉํ•˜์ง€๋งŒ ์–‘์žํ™” ์ •ํ™•๋„๊ฐ€ ๋‚ฎ์•„์ง‘๋‹ˆ๋‹ค. "None"์€ ๊ฐ€๋Šฅํ•œ ๊ฐ€์žฅ ๋‚ฎ์€ ๊ฐ’์ž…๋‹ˆ๋‹ค.
45
+ - Act Order: True or False. `desc_act`์œผ๋กœ๋„ ์•Œ๋ ค์ ธ ์žˆ์Šต๋‹ˆ๋‹ค. ์ฐธ์ด๋ฉด ์–‘์žํ™” ์ •ํ™•๋„๊ฐ€ ํ–ฅ์ƒ๋ฉ๋‹ˆ๋‹ค.
46
+ - Damp %: ์ƒ˜ํ”Œ์ด ์ •๋Ÿ‰ํ™”๋ฅผ ์œ„ํ•ด ์ฒ˜๋ฆฌ๋˜๋Š” ๋ฐฉ์‹์— ์˜ํ–ฅ์„ ์ฃผ๋Š” GPTQ ๋งค๊ฐœ๋ณ€์ˆ˜์ž…๋‹ˆ๋‹ค. 0.01์ด ๊ธฐ๋ณธ๊ฐ’์ด์ง€๋งŒ 0.1์„ ์‚ฌ์šฉํ•˜๋ฉด ์ •ํ™•๋„๊ฐ€ ์•ฝ๊ฐ„ ํ–ฅ์ƒ๋ฉ๋‹ˆ๋‹ค.
47
+ - GPTQ dataset: ์ •๋Ÿ‰ํ™”์— ์‚ฌ์šฉ๋˜๋Š” ๋ฐ์ดํ„ฐ ์„ธํŠธ์ž…๋‹ˆ๋‹ค. ๋ชจ๋ธ ํ•™์Šต์— ๋” ์ ํ•ฉํ•œ ๋ฐ์ดํ„ฐ ์„ธํŠธ๋ฅผ ์‚ฌ์šฉํ•˜๋ฉด ์ •๋Ÿ‰ํ™” ์ •ํ™•๋„๊ฐ€ ํ–ฅ์ƒ๋  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. GPTQ ๋ฐ์ดํ„ฐ ์„ธํŠธ๋Š” ๋ชจ๋ธ ํ•™์Šต์— ์‚ฌ์šฉ๋œ ๋ฐ์ดํ„ฐ ์„ธํŠธ์™€ ๋™์ผํ•˜์ง€ ์•Š์œผ๋ฏ€๋กœ ํ•™์Šต ๋ฐ์ดํ„ฐ ์„ธํŠธ์— ๋Œ€ํ•œ ์ž์„ธํ•œ ๋‚ด์šฉ์€ ์›๋ณธ ๋ชจ๋ธ repo๋ฅผ ์ฐธ์กฐํ•˜์„ธ์š”.
48
+ - Sequence Length: ์ •๋Ÿ‰ํ™”์— ์‚ฌ์šฉ๋œ ๋ฐ์ดํ„ฐ ์„ธํŠธ ์‹œํ€€์Šค์˜ ๊ธธ์ด์ž…๋‹ˆ๋‹ค. ์ด์ƒ์ ์œผ๋กœ๋Š” ๋ชจ๋ธ ์‹œํ€€์Šค ๊ธธ์ด์™€ ๋™์ผํ•ฉ๋‹ˆ๋‹ค. ์ผ๋ถ€ ๋งค์šฐ ๊ธด ์‹œํ€€์Šค ๋ชจ๋ธ(16+K)์˜ ๊ฒฝ์šฐ ๋” ์งง์€ ์‹œํ€€์Šค ๊ธธ์ด๋ฅผ ์‚ฌ์šฉํ•ด์•ผ ํ•  ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค. ์‹œํ€€์Šค ๊ธธ์ด๊ฐ€ ์งง๋‹ค๊ณ  ํ•ด์„œ ์–‘์žํ™”๋œ ๋ชจ๋ธ์˜ ์‹œํ€€์Šค ๊ธธ์ด๊ฐ€ ์ œํ•œ๋˜๋Š” ๊ฒƒ์€ ์•„๋‹™๋‹ˆ๋‹ค. ์ด๋Š” ๊ธด ์ถ”๋ก  ์‹œํ€€์Šค์˜ ์–‘์žํ™” ์ •ํ™•๋„์—๋งŒ ์˜ํ–ฅ์„ ๋ฏธ์นฉ๋‹ˆ๋‹ค.
49
+ - ExLlama Compatibility: Exllama๋กœ ์ด ํŒŒ์ผ์„ ๋กœ๋“œํ•  ์ˆ˜ ์žˆ๋Š”์ง€์˜ ์—ฌ๋ถ€์ด๋ฉฐ, ํ˜„์žฌ 4๋น„ํŠธ์˜ ๋ผ๋งˆ ๋ชจ๋ธ๋งŒ ์ง€์›ํ•ฉ๋‹ˆ๋‹ค.
50
 
51
  </details>
52
 
53
+ | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc |
54
+ | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- |
55
+ | [main](https://huggingface.co/kuotient/llama-2-ko-70b-GPTQ/tree/main) | 4 | None | Yes | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 35.8 GB | Yes | 4-bit, Act Order ํฌํ•จ. VRAM ์‚ฌ์šฉ๋Ÿ‰์„ ์ค„์ด๊ธฐ ์œ„ํ•œ group size -1. |
56
+
57
+ <!-- README_GPTQ.md-provided-files end -->
58
+
59
+ <!-- original model card start -->
60
+
61
  # Original model card: Llama 2 ko 70b
62
  > ๐Ÿšง Note: this repo is under construction ๐Ÿšง
63
 
 
187
  publisher = { Hugging Face }
188
  }
189
  ```
190
+ <!-- original model card end -->