|
--- |
|
inference: false |
|
license: llama2 |
|
model_creator: WizardLM |
|
model_link: https://huggingface.co/WizardLM/WizardLM-70B-V1.0 |
|
model_name: WizardLM 70B V1.0 |
|
model_type: llama |
|
quantized_by: Thireus |
|
--- |
|
|
|
# WizardLM 70B V1.0 – EXL2 |
|
- Model creator: [WizardLM](https://huggingface.co/WizardLM) |
|
- FP32 Original model used for quantization: [WizardLM 70B V1.0](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) – float32 |
|
- FP16 Model used for quantization: [WizardLM 70B V1.0-HF](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF) – float16 of [WizardLM 70B V1.0](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) |
|
- BF16 Model used for quantization: [WizardLM 70B V1.0-BF16](https://huggingface.co/Thireus/WizardLM-70B-V1.0-BF16) – bfloat16 of [WizardLM 70B V1.0](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) |
|
|
|
## Models available: |
|
|
|
| Link | BITS (-b) | HEAD BITS (-hb) | MEASU-REMENT LENGTH (-ml) | LENGTH (-l) | CAL DATASET (-c) | Size | V. | Max Context Length | Base Model | Layers | VRAM Min*** | VRAM Max*** | PPL** | Comments                                                                                                                         | |
|
| ------ | --------- | --------------- | ------------------------ | ----------- | ---------------- | ---- | ------- | ------------------ | ---- | ---- |------------------ | ------------------ | ------------------ | ---------------------------------------------------------------------------------- | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-FP32-4.0bpw-h6-exl2/) | 4.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 33GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/c0dd3412d59c0bc776264512bf76264e954c221d) | 4096 | [FP32](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) | 80 | 39GB | 44GB | 4.15234375 | Good results | | [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-HF-4.0bpw-h6-exl2/) | 4.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 35GB | [0.0.1](https://github.com/turboderp/exllamav2/tree/aee7a281708d5faff2ad0ea4b3a3a4b754f458f3) | 4096 | [FP16](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF) | 80 | 40GB | 44GB | 4.1640625 | Model suffers from poor prompt understanding and logic is affected | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-BF16-4.0bpw-h6-exl2/) | 4.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 33GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/ec5164b8a8e282b91aedb2af94dfeb89887656b7) | 4096 | [BF16](https://huggingface.co/Thireus/WizardLM-70B-V1.0-BF16) | 80 | 39GB | 44GB | 4.2421875 | Model suffers from poor prompt understanding and logic is affected | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-HF-4.0bpw-h8-exl2/) | 4.0 | 8 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 35GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/a4f2663e310919f007c593030d56ca110f99c261) | 4096 | [FP16](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF) | 80 | 39GB | 44GB | 4.24609375 | Model suffers from poor prompt understanding and logic is affected | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-FP32-5.0bpw-h6-exl2/) | 5.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 41GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/c0dd3412d59c0bc776264512bf76264e954c221d) | 4096 | [FP32](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) | 80 | 47GB | 52GB | 4.06640625 | Best so far. Good results | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-HF-5.0bpw-h8-exl2/) | 5.0 | 8 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 44GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/a4f2663e310919f007c593030d56ca110f99c261) | 4096 | [FP16](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF) | 80 | 48GB | 52GB | 4.09765625 | Model suffers from poor prompt understanding and logic is affected | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-HF-5.0bpw-h6-exl2/) | 5.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 44GB | [0.0.1](https://github.com/turboderp/exllamav2/tree/aee7a281708d5faff2ad0ea4b3a3a4b754f458f3) | 4096 | [FP16](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF) | 80 | 48GB | 52GB | 4.0625 | Model suffers from poor prompt understanding and logic is affected | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-BF16-5.0bpw-h6-exl2/) | 5.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 41GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/ec5164b8a8e282b91aedb2af94dfeb89887656b7) | 4096 | [BF16](https://huggingface.co/Thireus/WizardLM-70B-V1.0-BF16) | 80 | 47GB | 52GB | 4.09765625 | Model suffers from poor prompt understanding and logic is affected | |
|
| [here](https://huggingface.co/Thireus/WizardLM-70B-V1.0-HF-6.0bpw-h6-exl2/) | 6.0 | 6 | 2048 | 2048 | [0000.parquet](https://huggingface.co/datasets/wikitext/tree/refs%2Fconvert%2Fparquet/wikitext-2-raw-v1/train)* | 49GB | [0.0.2](https://github.com/turboderp/exllamav2/tree/fae6fb296c6db4e3b1314c49c030541bed98acb9) | 4096 | [FP16](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF) | 80 | 56GB | 60GB | 4.0703125 | Model suffers from poor prompt understanding and logic is affected | |
|
|
|
|
|
\* wikitext-2-raw-v1 |
|
|
|
\*\* Evaluated with text-generation-webui ExLlama v0.0.2 on wikitext-2-raw-v1 (stride 512 and max_length 0). For reference, [TheBloke_WizardLM-70B-V1.0-GPTQ_gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/WizardLM-70B-V1.0-GPTQ/tree/gptq-4bit-32g-actorder_True) has a score of 4.1015625 in perplexity. |
|
|
|
\*\*\* Without Flash Attention - For better VRAM optimisation, make sure you install https://github.com/Dao-AILab/flash-attention#installation-and-features |
|
|
|
## Description: |
|
|
|
_This repository contains EXL2 model files for [WizardLM's WizardLM 70B V1.0](https://huggingface.co/WizardLM/WizardLM-70B-V1.0)._ |
|
|
|
EXL2 is a new format used by ExLlamaV2 – https://github.com/turboderp/exllamav2. EXL2 is based on the same optimization method as GPTQ. The format allows for mixing quantization |
|
levels within a model to achieve any average bitrate between 2 and 8 bits per weight. |
|
|
|
## Prompt template (official): |
|
|
|
``` |
|
A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT: |
|
``` |
|
|
|
## Prompt template (suggested): |
|
|
|
``` |
|
A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. |
|
USER: |
|
{prompt} |
|
ASSISTANT: |
|
|
|
|
|
``` |
|
|
|
## Quantization process: |
|
|
|
| Original Model | → | (optional) float16 or bfloat16 Model* | → | Safetensors Model** | → | EXL2 Model | |
|
| -------------- | --- | ------------- | --- | ---------------- | --- | ---------- | |
|
| [WizardLM 70B V1.0](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) | → | [WizardLM 70B V1.0-HF](https://huggingface.co/simsim314/WizardLM-70B-V1.0-HF)* | → | Safetensors** | → | EXL2 | |
|
|
|
Example to convert WizardLM-70B-V1.0-HF to EXL2 4.0 bpw with 6-bit head: |
|
|
|
``` |
|
mkdir -p ~/EXL2/WizardLM-70B-V1.0-HF_4bit # Create the output directory |
|
python convert.py -i ~/float16_safetensored/WizardLM-70B-V1.0-HF -o ~/EXL2/WizardLM-70B-V1.0-HF_4bit -c ~/EXL2/0000.parquet -b 4.0 -hb 6 |
|
``` |
|
|
|
\* Use the following script to convert your local pytorch_model bin files to float16 (you can also choose bfloat16) + safetensors all in one go: |
|
|
|
- https://github.com/oobabooga/text-generation-webui/blob/main/convert-to-safetensors.py |
|
(best for sharding and float16/FP16 or bfloat16/BF16 conversion) |
|
|
|
Example to convert [WizardLM 70B V1.0](https://huggingface.co/WizardLM/WizardLM-70B-V1.0) directly to float16 safetensors in 10GB shards: |
|
|
|
``` |
|
python convert-to-safetensors.py ~/original/WizardLM-70B-V1.0 --output ~/float16_safetensored/WizardLM-70B-V1.0 --max-shard-size 10GB |
|
``` |
|
|
|
Use `--bf16` if you'd like to try bfloat16 instead, but note that there are concerns about quantization quality – https://github.com/turboderp/exllamav2/issues/30#issuecomment-1719009289 |
|
|
|
\*\* Use any one of the following scripts to convert your local pytorch_model bin files to safetensors: |
|
|
|
- https://github.com/turboderp/exllamav2/blob/master/util/convert_safetensors.py (official ExLlamaV2) |
|
- https://huggingface.co/Panchovix/airoboros-l2-70b-gpt4-1.4.1-safetensors/blob/main/bin2safetensors/convert.py (recommended) |
|
- https://gist.github.com/epicfilemcnulty/1f55fd96b08f8d4d6693293e37b4c55e#file-2safetensors-py |
|
|
|
## Further reading: |
|
|
|
- https://mlabonne.github.io/blog/posts/Introduction_to_Weight_Quantization.html |
