jpacifico commited on
Commit
88f25e5
1 Parent(s): ca46218

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +28 -23
README.md CHANGED
@@ -101,6 +101,34 @@ French-Alpaca-7B-Instruct_beta 5.587866
101
  vigogne-2-7b-chat 4.218750
102
  ```
103
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
104
  ### Usage
105
 
106
  You can run this model using my [Colab notebook](https://github.com/jpacifico/Chocolatine-LLM/blob/main/Chocolatine_3B_inference_test_colab.ipynb)
@@ -138,29 +166,6 @@ sequences = pipeline(
138
  print(sequences[0]['generated_text'])
139
  ```
140
 
141
- * **4-bit quantized version** is available here : [jpacifico/Chocolatine-3B-Instruct-DPO-Revised-Q4_K_M-GGUF](https://huggingface.co/jpacifico/Chocolatine-3B-Instruct-DPO-Revised-Q4_K_M-GGUF)
142
-
143
- * **Ollama**: [jpacifico/chocolatine-3b](https://ollama.com/jpacifico/chocolatine-3b)
144
-
145
- ```bash
146
- ollama run jpacifico/chocolatine-3b
147
- ```
148
-
149
- Ollama *Modelfile* example :
150
-
151
- ```bash
152
- FROM ./chocolatine-3b-instruct-dpo-revised-q4_k_m.gguf
153
- TEMPLATE """{{ if .System }}<|system|>
154
- {{ .System }}<|end|>
155
- {{ end }}{{ if .Prompt }}<|user|>
156
- {{ .Prompt }}<|end|>
157
- {{ end }}<|assistant|>
158
- {{ .Response }}<|end|>
159
- """
160
- PARAMETER stop """{"stop": ["<|end|>","<|user|>","<|assistant|>"]}"""
161
- SYSTEM """You are a friendly assistant called Chocolatine."""
162
- ```
163
-
164
  ### Limitations
165
 
166
  The Chocolatine model is a quick demonstration that a base model can be easily fine-tuned to achieve compelling performance.
 
101
  vigogne-2-7b-chat 4.218750
102
  ```
103
 
104
+
105
+ ### Quantized versions
106
+
107
+ * **4-bit quantized version** is available here : [jpacifico/Chocolatine-3B-Instruct-DPO-Revised-Q4_K_M-GGUF](https://huggingface.co/jpacifico/Chocolatine-3B-Instruct-DPO-Revised-Q4_K_M-GGUF)
108
+
109
+ * **8-bit quantized version** also available here : [jpacifico/Chocolatine-3B-Instruct-DPO-Revised-Q8_0-GGUF](https://huggingface.co/jpacifico/Chocolatine-3B-Instruct-DPO-Revised-Q8_0-GGUF)
110
+
111
+ * **Ollama**: [jpacifico/chocolatine-3b](https://ollama.com/jpacifico/chocolatine-3b)
112
+
113
+ ```bash
114
+ ollama run jpacifico/chocolatine-3b
115
+ ```
116
+
117
+ Ollama *Modelfile* example :
118
+
119
+ ```bash
120
+ FROM ./chocolatine-3b-instruct-dpo-revised-q4_k_m.gguf
121
+ TEMPLATE """{{ if .System }}<|system|>
122
+ {{ .System }}<|end|>
123
+ {{ end }}{{ if .Prompt }}<|user|>
124
+ {{ .Prompt }}<|end|>
125
+ {{ end }}<|assistant|>
126
+ {{ .Response }}<|end|>
127
+ """
128
+ PARAMETER stop """{"stop": ["<|end|>","<|user|>","<|assistant|>"]}"""
129
+ SYSTEM """You are a friendly assistant called Chocolatine."""
130
+ ```
131
+
132
  ### Usage
133
 
134
  You can run this model using my [Colab notebook](https://github.com/jpacifico/Chocolatine-LLM/blob/main/Chocolatine_3B_inference_test_colab.ipynb)
 
166
  print(sequences[0]['generated_text'])
167
  ```
168
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
169
  ### Limitations
170
 
171
  The Chocolatine model is a quick demonstration that a base model can be easily fine-tuned to achieve compelling performance.