Update README.md
Browse files
README.md
CHANGED
|
@@ -32,12 +32,7 @@ This repository contains a **4-bit quantized version** of the MiniMax-M2 model.
|
|
| 32 |
The quantization was performed using **[GPTQModel](https://github.com/ModelCloud/GPTQModel)** with an experimental modification that **feeds the whole dataset to each expert** to achieve improved quality.
|
| 33 |
|
| 34 |
**Calibration Dataset:**
|
| 35 |
-
The dataset used during quantization consists of 1536 samples:
|
| 36 |
-
* c4/en (1024)
|
| 37 |
-
* arc (168)
|
| 38 |
-
* gsm8k (168)
|
| 39 |
-
* humaneval (168)
|
| 40 |
-
* alpaca (20)
|
| 41 |
|
| 42 |
**Hardware & Performance:**
|
| 43 |
This model is verified to run with Tensor Parallel (TP) on **8x NVIDIA RTX 3090** GPUs with a context window of **192,500 tokens**.
|
|
|
|
| 32 |
The quantization was performed using **[GPTQModel](https://github.com/ModelCloud/GPTQModel)** with an experimental modification that **feeds the whole dataset to each expert** to achieve improved quality.
|
| 33 |
|
| 34 |
**Calibration Dataset:**
|
| 35 |
+
The dataset used during quantization consists of 1536 samples: c4/en (1024), arc (164), gsm8k (164), humaneval (164), alpaca (20)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
|
| 37 |
**Hardware & Performance:**
|
| 38 |
This model is verified to run with Tensor Parallel (TP) on **8x NVIDIA RTX 3090** GPUs with a context window of **192,500 tokens**.
|