measurement.json

README.md

@@ -0,0 +1,59 @@
+---
+pipeline_tag: text-generation
+license: other
+quantized_by: bartowski
+---
+
+## Exllama v2 Quantizations of internlm2-7b-llama
+
+Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.11">turboderp's ExLlamaV2 v0.0.11</a> for quantization.
+
+# The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)
+
+Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
+
+Conversion was done using the default calibration dataset.
+
+Default arguments used except when the bits per weight is above 6.0, at that point the lm_head layer is quantized at 8 bits per weight instead of the default 6.
+
+Original model: https://huggingface.co/internlm/internlm2-7b
+
+Model Size: 7b
+
+| Branch | Bits | lm_head | Dataset | Size | Description |
+| ----- | ---- | ------- | ------- | ------ | ------------ |
+| [8_0](https://huggingface.co/Bartowski/internlm2-7b-llama-exl2/tree/8_0) | 8.0  | 8.0 | Default | 9.8 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
+| [6_5](https://huggingface.co/Bartowski/internlm2-7b-llama-exl2/tree/6_5) | 6.5  | 8.0 | Default | 8.6 GB | Very similar to 8.0, good tradeoff of size vs performance, **recommended**. |
+| [5_0](https://huggingface.co/Bartowski/internlm2-7b-llama-exl2/tree/5_0) | 5.0  | 6.0 | Default | 7.4 GB | Slightly lower perplexity vs 6.5. |
+| [4_25](https://huggingface.co/Bartowski/internlm2-7b-llama-exl2/tree/4_25) | 4.25 | 6.0 | Default | 6.7 GB | GPTQ equivalent bits per weight. |
+| [3_5](https://huggingface.co/Bartowski/internlm2-7b-llama-exl2/tree/3_5) | 3.5  | 6.0 | Default | 6.1 GB | Lower quality, only use if you have to. |
+
+All VRAM requirements estimated from 16k context. For 32k context add ~2 GB.
+
+## Download instructions
+
+With git:
+
+```shell
+git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/internlm2-7b-llama-exl2 internlm2-7b-llama-exl2-6_5
+```
+
+With huggingface hub (credit to TheBloke for instructions):
+
+```shell
+pip3 install huggingface-hub
+```
+
+To download the `main` (only useful if you only care about measurement.json) branch to a folder called `internlm2-7b-llama-exl2`:
+
+```shell
+mkdir internlm2-7b-llama-exl2
+huggingface-cli download bartowski/internlm2-7b-llama-exl2 --local-dir internlm2-7b-llama-exl2 --local-dir-use-symlinks False
+```
+
+To download from a different branch, add the `--revision` parameter:
+
+```shell
+mkdir internlm2-7b-llama-exl2-6_5
+huggingface-cli download bartowski/internlm2-7b-llama-exl2 --revision 6_5 --local-dir internlm2-7b-llama-exl2-6_5 --local-dir-use-symlinks False
+```