alpayariyak commited on
Commit
29b3d50
β€’
1 Parent(s): 96cc489

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +39 -15
README.md CHANGED
@@ -1,3 +1,4 @@
 
1
  <div align="center">
2
  <img src="https://raw.githubusercontent.com/imoneoi/openchat/master/assets/logo_new.png" style="width: 65%">
3
  <h1>Advancing Open-source Language Models with Mixed-Quality Data</h1>
@@ -23,31 +24,48 @@
23
  </p>
24
 
25
  <hr>
26
- <div style="background-color: white; padding: 0.7em; border-radius: 0.5em; color: black; display: flex; flex-direction: column; justify-content: center; text-align: center; ont-size: 0.5em;">
27
  <a href="https://huggingface.co/openchat/openchat_3.5" style="text-decoration: none; color: black;">
28
- <span style="font-size: 0.7em; font-family: 'Helvetica'; color: white; vertical-align: top; background-color:white; border-radius: 6em; padding: 0.04em 0.4em; letter-spacing: 0.1em; font-weight: bold">3.51210</span>
29
- <span style="font-size: 1.7em; font-family: 'Helvetica'; letter-spacing: 0.1em; font-weight: bold; color: black;">OPENCHAT</span><span style="font-size: 1.8em; font-family: 'Helvetica'; color: #3c72db; ">3.5</span>
30
- <span style="font-size: 0.7em; font-family: 'Helvetica'; color: white; vertical-align: top; background-color:red; border-radius: 6em; padding: 0.066em 0.4em; letter-spacing: 0.1em; font-weight: bold;">1210</span>
31
- <span style="font-size: 1em; font-family: 'Helvetica'; color: black;">
32
- <br> πŸ† The Overall Best Performing Open Source 7B Model πŸ†
33
- <br> πŸ€– Outperforms <span style="font-weight: bold;">ChatGPT</span> (March) and <span style="font-weight: bold;">Grok-1</span> on most benchmarks πŸ€–
34
- <br> πŸš€<span style="font-size: 1em; font-family: 'Helvetica'; color: black; font-weight: bold;">15</span>-point improvement in Coding Performance over <span style="font-size: 0.9em;
35
- font-family: 'Helvetica'; color: black; font-weight: bold;">OpenChat-3.5πŸš€</span>
36
- <br><span style="font-size: 1em; font-family: 'Helvetica'; color: #3c72db; font-weight: bold;">New Features</span>
37
- <br> πŸ’‘ 2 Modes: Coding + Generalist, Mathematical Reasoning πŸ’‘
38
- <br> πŸ§‘β€βš–οΈ Experimental support for Evaluator and Feedback capabilities πŸ§‘β€βš–οΈ
 
39
  </span>
40
  </a>
41
  </div>
42
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
43
  <div style="display: flex; justify-content: center; align-items: center">
44
- <img src="https://github.com/alpayariyak/openchat/blob/master/assets/1210bench.png?raw=true" style="width: 100%; border-radius: 1em">
45
  </div>
46
 
47
  <div>
48
  <h3> Table of Contents</h3>
49
  </div>
50
 
 
51
  1. [Usage](#usage)
52
  2. [Benchmarks](#benchmarks)
53
  3. [Limitations](#limitations)
@@ -174,7 +192,6 @@ Score 5: {orig_score5_description}
174
  | OpenOrca Mistral | 7B | 52.7 | 6.86 | 38.4 | 49.4 | 42.9 | 45.9 | 59.3 | 59.1 | 58.1 |
175
  | Zephyr-Ξ²^ | 7B | 34.6 | 7.34 | 22.0 | 40.6 | 39.0 | 40.8 | 39.8 | 5.1 | 16.0 |
176
  | Mistral | 7B | - | 6.84 | 30.5 | 39.0 | 38.0 | - | 60.1 | 52.2 | - |
177
-
178
  <details>
179
  <summary>Evaluation Details(click to expand)</summary>
180
  *: ChatGPT (March) results are from [GPT-4 Technical Report](https://arxiv.org/abs/2303.08774), [Chain-of-Thought Hub](https://github.com/FranxYao/chain-of-thought-hub), and our evaluation. Please note that ChatGPT is not a fixed baseline and evolves rapidly over time.
@@ -189,6 +206,7 @@ All models are evaluated in chat mode (e.g. with the respective conversation tem
189
  <h3>HumanEval+</h3>
190
  </div>
191
 
 
192
  | Model | Size | HumanEval+ pass@1 |
193
  |-----------------------------|----------|------------|
194
  | ChatGPT (December 12, 2023) | - | 64.6 |
@@ -209,6 +227,12 @@ All models are evaluated in chat mode (e.g. with the respective conversation tem
209
 
210
  *: Grok results are reported by [X.AI](https://x.ai/).
211
 
 
 
 
 
 
 
212
  <div align="center">
213
  <h2> Limitations </h2>
214
  </div>
@@ -226,6 +250,7 @@ OpenChat may sometimes generate information that does not exist or is not accura
226
  **Safety**
227
  OpenChat may sometimes generate harmful, hate speech, biased responses, or answer unsafe questions. It's crucial to apply additional AI safety measures in use cases that require safe and moderated responses.
228
 
 
229
  <div align="center">
230
  <h2> License </h2>
231
  </div>
@@ -251,7 +276,6 @@ OpenChat 3.5 was trained with C-RLFT on a collection of publicly available high-
251
  <div align="center">
252
  <h2> Citation </h2>
253
  </div>
254
-
255
  ```
256
  @article{wang2023openchat,
257
  title={OpenChat: Advancing Open-source Language Models with Mixed-Quality Data},
 
1
+
2
  <div align="center">
3
  <img src="https://raw.githubusercontent.com/imoneoi/openchat/master/assets/logo_new.png" style="width: 65%">
4
  <h1>Advancing Open-source Language Models with Mixed-Quality Data</h1>
 
24
  </p>
25
 
26
  <hr>
27
+ <div style="background-color: white; padding: 0.7em; border-radius: 0.5em; color: black; display: flex; flex-direction: column; justify-content: center; text-align: center;">
28
  <a href="https://huggingface.co/openchat/openchat_3.5" style="text-decoration: none; color: black;">
29
+ <span style="font-size: 0.7em; font-family: 'Helvetica'; color: white; background-color:white; border-radius: 6em; padding: 0.04em 0.4em; letter-spacing: 0.1em; font-weight: bold">3.51210</span>
30
+ <span style="font-size: 1.7em; font-family: 'Helvetica'; letter-spacing: 0.1em; font-weight: bold; color: black;">OPENCHAT</span><span style="font-size: 1.8em; font-family: 'Helvetica'; color: #3c72db;">3.5</span>
31
+ <span style="font-size: 0.7em; font-family: 'Helvetica'; color: white; background-color:red; border-radius: 6em; padding: 0.066em 0.4em; letter-spacing: 0.1em; font-weight: bold; vertical-align: top;">1210</span><br>
32
+ <span style="font-size: 2vw; font-family: 'Helvetica'; color: black; white-space: nowrap;">
33
+ πŸ† The Overall Best Performing Open Source 7B Model πŸ†
34
+ </span>
35
+ <br> <span style="font-size: 2vw; font-family: 'Helvetica'; color: black; white-space: nowrap;">πŸ€– Outperforms <span style="font-weight: bold;">ChatGPT</span> (March) and <span style="font-weight: bold;">Grok-1</span> on most benchmarks πŸ€–</span>
36
+ <br> <span style="font-size: 2vw; font-family: 'Helvetica'; color: black; white-space: nowrap;">πŸš€ <span style="font-size: 1em; font-family: 'Helvetica'; color: black; font-weight: bold;">15</span>-point improvement in Coding Performance over <span style="font-size: 0.9em;
37
+ font-family: 'Helvetica'; color: black; font-weight: bold;">OpenChat-3.5 πŸš€</span></span>
38
+ <br><span style="font-size: 2vw; font-family: 'Helvetica'; color: #3c72db; font-weight: bold; white-space: nowrap;">New Features</span>
39
+ <br> <span style="font-size: 2vw; font-family: 'Helvetica'; color: black; white-space: nowrap;">πŸ’‘ 2 Modes: Coding + Generalist, Mathematical Reasoning πŸ’‘</span>
40
+ <br><span style="font-size: 2vw; font-family: 'Helvetica'; color: black; white-space: nowrap;"> πŸ§‘β€βš–οΈ Experimental support for Evaluator and Feedback capabilities πŸ§‘β€βš–οΈ</span>
41
  </span>
42
  </a>
43
  </div>
44
 
45
+ <!-- <a href="https://huggingface.co/openchat/openchat_3.5">
46
+ <button class="common-button">Model Repo</button>
47
+ </a>
48
+ <a href="https://openchat.team">
49
+ <button class="common-button">OpenChatUI Demo</button>
50
+ </a>
51
+ <a href="https://huggingface.co/spaces/openchat/openchat_3.5">
52
+ <button class="common-button">HuggingFace Space</button>
53
+ </a>
54
+ <a href="https://arxiv.org/pdf/2309.11235.pdf">
55
+ <button class="common-button">Paper</button>
56
+ </a>
57
+ -->
58
+ </p>
59
+
60
  <div style="display: flex; justify-content: center; align-items: center">
61
+ <img src="https://github.com/alpayariyak/openchat/blob/master/assets/1210bench.png?raw=true" style="width: 100%; border-radius: 1em">">
62
  </div>
63
 
64
  <div>
65
  <h3> Table of Contents</h3>
66
  </div>
67
 
68
+
69
  1. [Usage](#usage)
70
  2. [Benchmarks](#benchmarks)
71
  3. [Limitations](#limitations)
 
192
  | OpenOrca Mistral | 7B | 52.7 | 6.86 | 38.4 | 49.4 | 42.9 | 45.9 | 59.3 | 59.1 | 58.1 |
193
  | Zephyr-Ξ²^ | 7B | 34.6 | 7.34 | 22.0 | 40.6 | 39.0 | 40.8 | 39.8 | 5.1 | 16.0 |
194
  | Mistral | 7B | - | 6.84 | 30.5 | 39.0 | 38.0 | - | 60.1 | 52.2 | - |
 
195
  <details>
196
  <summary>Evaluation Details(click to expand)</summary>
197
  *: ChatGPT (March) results are from [GPT-4 Technical Report](https://arxiv.org/abs/2303.08774), [Chain-of-Thought Hub](https://github.com/FranxYao/chain-of-thought-hub), and our evaluation. Please note that ChatGPT is not a fixed baseline and evolves rapidly over time.
 
206
  <h3>HumanEval+</h3>
207
  </div>
208
 
209
+
210
  | Model | Size | HumanEval+ pass@1 |
211
  |-----------------------------|----------|------------|
212
  | ChatGPT (December 12, 2023) | - | 64.6 |
 
227
 
228
  *: Grok results are reported by [X.AI](https://x.ai/).
229
 
230
+
231
+
232
+
233
+
234
+
235
+
236
  <div align="center">
237
  <h2> Limitations </h2>
238
  </div>
 
250
  **Safety**
251
  OpenChat may sometimes generate harmful, hate speech, biased responses, or answer unsafe questions. It's crucial to apply additional AI safety measures in use cases that require safe and moderated responses.
252
 
253
+ ## License
254
  <div align="center">
255
  <h2> License </h2>
256
  </div>
 
276
  <div align="center">
277
  <h2> Citation </h2>
278
  </div>
 
279
  ```
280
  @article{wang2023openchat,
281
  title={OpenChat: Advancing Open-source Language Models with Mixed-Quality Data},