Safetensors
qwen2
weinz commited on
Commit
7705b60
1 Parent(s): 1808d21

Update README.md

Browse files

edit the model card

Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -19,10 +19,10 @@ We're excited to introduce Hammer 2.0, the latest in our Hammer Large Language M
19
  Hammer2.0 finetuned based on [Qwen 2.5 series](https://huggingface.co/collections/Qwen/qwen25-66e81a666513e518adb90d9e) and [Qwen 2.5 coder series](https://huggingface.co/collections/Qwen/qwen25-coder-66eaa22e6f99801bf65b0c2f). It's trained using the [APIGen Function Calling Datasets](https://huggingface.co/datasets/Salesforce/xlam-function-calling-60k) containing 60,000 samples, supplemented by [7,500 irrelevance detection data](https://huggingface.co/datasets/MadeAgents/XLAM-7.5k-Irrelevance) we generated. Employing innovative training techniques of function masking, function shuffling, and prompt optimization, Hammer2.0 has achieved exceptional performances across numerous benchmarks including [Berkley Function Calling Leaderboard](https://gorilla.cs.berkeley.edu/leaderboard.html), [API-Bank](https://arxiv.org/abs/2304.08244), [Tool-Alpaca](https://arxiv.org/abs/2306.05301), [Nexus Raven](https://github.com/nexusflowai/NexusRaven-V2) and [Seal-Tools](https://arxiv.org/abs/2405.08355).
20
 
21
  ## Tuning Details
22
- We will soon release a report detailing our models' technical aspects. Stay tuned!
23
 
24
  ## Evaluation
25
- The evaluation result of Hammer2.0 series on the Berkeley Function-Calling Leaderboard (BFCL) are presented in the following table:
26
  <div style="text-align: center;">
27
  <img src="v2_figures/bfcl.PNG" alt="overview" width="1000" style="margin: auto;">
28
  </div>
@@ -33,7 +33,7 @@ In addition, we evaluated Hammer2.0 on other academic benchmarks to further show
33
  <img src="v2_figures/others.PNG" alt="overview" width="1000" style="margin: auto;">
34
  </div>
35
 
36
- On comparison, Hammer 2.0 outperform models with similar sizes and even surpass many larger models overall.
37
 
38
  ## Requiements
39
  The code of Hammer2.0-7b has been in the latest Hugging face transformers and we advise you to install `transformers>=4.37.0`.
 
19
  Hammer2.0 finetuned based on [Qwen 2.5 series](https://huggingface.co/collections/Qwen/qwen25-66e81a666513e518adb90d9e) and [Qwen 2.5 coder series](https://huggingface.co/collections/Qwen/qwen25-coder-66eaa22e6f99801bf65b0c2f). It's trained using the [APIGen Function Calling Datasets](https://huggingface.co/datasets/Salesforce/xlam-function-calling-60k) containing 60,000 samples, supplemented by [7,500 irrelevance detection data](https://huggingface.co/datasets/MadeAgents/XLAM-7.5k-Irrelevance) we generated. Employing innovative training techniques of function masking, function shuffling, and prompt optimization, Hammer2.0 has achieved exceptional performances across numerous benchmarks including [Berkley Function Calling Leaderboard](https://gorilla.cs.berkeley.edu/leaderboard.html), [API-Bank](https://arxiv.org/abs/2304.08244), [Tool-Alpaca](https://arxiv.org/abs/2306.05301), [Nexus Raven](https://github.com/nexusflowai/NexusRaven-V2) and [Seal-Tools](https://arxiv.org/abs/2405.08355).
20
 
21
  ## Tuning Details
22
+ We will soon release a report detailing the technical aspects of our models. Stay tuned!
23
 
24
  ## Evaluation
25
+ The evaluation results of Hammer 2.0 series on the Berkeley Function-Calling Leaderboard (BFCL) are presented in the following table:
26
  <div style="text-align: center;">
27
  <img src="v2_figures/bfcl.PNG" alt="overview" width="1000" style="margin: auto;">
28
  </div>
 
33
  <img src="v2_figures/others.PNG" alt="overview" width="1000" style="margin: auto;">
34
  </div>
35
 
36
+ On comparison, Hammer 2.0 outperforms models with similar sizes and even surpass many larger models overall.
37
 
38
  ## Requiements
39
  The code of Hammer2.0-7b has been in the latest Hugging face transformers and we advise you to install `transformers>=4.37.0`.