Upload README.md with huggingface_hub

README.md

@@ -1,14 +1,11 @@
 ---
-base_model: aws-prototyping/MegaBeam-Mistral-7B-512k
-license: apache-2.0
-pipeline_tag: text-generation
 quantized_by: bartowski
-inference: false
+pipeline_tag: text-generation
 ---
 
 ## Llamacpp imatrix Quantizations of MegaBeam-Mistral-7B-512k
 
-Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggerganov/llama.cpp/releases/tag/b3496">b3496</a> for quantization.
+Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggerganov/llama.cpp/releases/tag/b3600">b3600</a> for quantization.
 
 Original model: https://huggingface.co/aws-prototyping/MegaBeam-Mistral-7B-512k
 
@@ -19,16 +16,20 @@ Run them in [LM Studio](https://lmstudio.ai/)
 ## Prompt format
 
 ```
-<s> [INST] {prompt} [/INST]</s>
+<s> [INST] {prompt} [/INST]
 ```
 
 Note that this model does not support a System prompt.
 
+## What's new:
+
+Model updated for "improved user experience" and fixing repetition issues
+
 ## Download a file (not the whole branch) from below:
 
 | Filename | Quant type | File Size | Split | Description |
 | -------- | ---------- | --------- | ----- | ----------- |
-| [MegaBeam-Mistral-7B-512k-f32.gguf](https://huggingface.co/bartowski/MegaBeam-Mistral-7B-512k-GGUF/blob/main/MegaBeam-Mistral-7B-512k-f32.gguf) | f32 | 28.97GB | false | Full F32 weights. |
+| [MegaBeam-Mistral-7B-512k-f16.gguf](https://huggingface.co/bartowski/MegaBeam-Mistral-7B-512k-GGUF/blob/main/MegaBeam-Mistral-7B-512k-f16.gguf) | f16 | 14.48GB | false | Full F16 weights. |
 | [MegaBeam-Mistral-7B-512k-Q8_0.gguf](https://huggingface.co/bartowski/MegaBeam-Mistral-7B-512k-GGUF/blob/main/MegaBeam-Mistral-7B-512k-Q8_0.gguf) | Q8_0 | 7.70GB | false | Extremely high quality, generally unneeded but max available quant. |
 | [MegaBeam-Mistral-7B-512k-Q6_K_L.gguf](https://huggingface.co/bartowski/MegaBeam-Mistral-7B-512k-GGUF/blob/main/MegaBeam-Mistral-7B-512k-Q6_K_L.gguf) | Q6_K_L | 6.01GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. |
 | [MegaBeam-Mistral-7B-512k-Q6_K.gguf](https://huggingface.co/bartowski/MegaBeam-Mistral-7B-512k-GGUF/blob/main/MegaBeam-Mistral-7B-512k-Q6_K.gguf) | Q6_K | 5.94GB | false | Very high quality, near perfect, *recommended*. |