Update README.md
Browse files
README.md
CHANGED
|
@@ -2,4 +2,10 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
|
| 5 |
-
4bit GGUF quantization of [TinyLlama-1.1B-intermediate-step-955k-token-2T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-955k-token-2T)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
|
| 5 |
+
4bit GGUF quantization of [TinyLlama-1.1B-intermediate-step-955k-token-2T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-955k-token-2T)
|
| 6 |
+
|
| 7 |
+
I used [this script] https://github.com/ggerganov/llama.cpp/blob/master/examples/make-ggml.py to generate the file using this command:
|
| 8 |
+
|
| 9 |
+
`python make-ggml.py ~/ooba/models/TinyLlama_TinyLlama-1.1B-intermediate-step-955k-token-2T/ --model_type=llama --quants=Q4_K_M`
|
| 10 |
+
|
| 11 |
+
The original model is so small that there is only one safetensors file named `model.safetensors`, so I had to change that filename to `model-00001-of-00001.safetensors` to make the script load the model properly.
|