Upload folder using huggingface_hub

Arch-Function-3B-Q2_K.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ac9a5a3d3ec8188b711c4c2601458cb11dd40ac47f435ddf2ac1adad06163dc2
+oid sha256:f66f014a1a9c6b74fa032e92f9cc0c1c3be42d5105df5b3f31b0d70a988608c7
 size 1274755488

Arch-Function-3B-Q3_K_L.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f3ebb8209e432f57e9d7513d3bb60cfb523414d5ec68d316afb5c765a97c9b92
+oid sha256:4be3cbad0651c5f148d8332bb570ad20bda79c7ba9e7b4da6ff85a9009bb6811
 size 1707391392

Arch-Function-3B-Q3_K_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:badb6a19390fe7abcc0325b435e6ddf7ed0cd30d46967987a896bfaefe3ee6e8
+oid sha256:f0a2d7ab7d524180b807563533841dd45ddf655c2cb3e667261af69ee187969e
 size 1590475168

Arch-Function-3B-Q3_K_S.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c71b8fdd436c7a2d1f00ed4322a28b977c292b3dd6a798c0400e40d0ed14c03
+oid sha256:066c42e06aa2b625a279e7fdf4f41d469e5734005590d4c74525c00307031762
 size 1454356896

Arch-Function-3B-Q4_K_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7172964e86c7ffa28c7931f1bfb846e4c9d304b74439f14c63de0e918d711f0b
+oid sha256:681134d4e73e20bcbd14e5870fdde27ea5e9f9ca98c7c11b5d8886387dfc0684
 size 1929902496

Arch-Function-3B-Q4_K_S.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3e80ad78953761dd86db7a7c2256ddd528b0c23f95a1e31be1c8bc93e3bf41fb
+oid sha256:6e638604b973dec2b54f38d2f31594cd37944fe77411c11e9bb266a5ea691b4e
 size 1834383776

Arch-Function-3B-Q5_K_M.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0dd4e6a39763d2175c0b3a3c64f96b2c391f9b5c3ebc01c1fa26493131173095
+oid sha256:e81a01993cb5e6e14786b7b2aa2003e50b8e6a8b40c8fe1d5ad7c95e65128907
 size 2224814496

Arch-Function-3B-Q5_K_S.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:294d4570d2cedc163c4e31d1385d1f8208bd8441b2fb281cdc58bc03c3c1f199
+oid sha256:9e7248944a554e0abe9b79fdbbe68c56a831c799d73171c6797b4022f0c653d3
 size 2169665952

Arch-Function-3B-Q6_K.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5b4ae2627be1de16a42e01103fbd399028394fc3e28bc2cecaf97eb84c8dded3
+oid sha256:62149b0a091bee66646f8f651f397aa8cf2aa092f39c280dd519e94af49b8991
 size 2538158496

Arch-Function-3B.gguf

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74f797d8c23e83f887de4d5cfd13db3d3de4bdb05bd5345b674e014f51f53c0b
+oid sha256:5db7b9f6d9a1570599307b90ec9271f228c736c688fab4083e31369842dd78f4
 size 6178316704

README.md

@@ -2,16 +2,16 @@
 license: other
 license_name: katanemo-research
 license_link: >-
-  https://huggingface.co/katanemolabs/Arch-Function-1.5B/blob/main/LICENSE
+  https://huggingface.co/katanemolabs/Arch-Function-3B.gguf/blob/main/LICENSE
 base_model:
-- Qwen/Qwen2.5-1.5B-Instruct
+- katanemo/Arch-Function-3B
 language:
 - en
 pipeline_tag: text-generation
 library_name: transformers
 ---
 
-# katanemo/Arch-Function-1.5B
+# katanemo/Arch-Function-3B
 
 ## Overview
 The Katanemo Arch-Function collection of large language models (LLMs) is a collection state-of-the-art (SOTA) LLMs specifically designed for **function calling** tasks. The models are designed to understand complex function signatures, identify required parameters, and produce accurate function call outputs based on natural language prompts. Achieving performance on par with GPT-4, these models set a new benchmark in the domain of function-oriented tasks, making them suitable for scenarios where automated API interaction and function execution is crucial.
@@ -54,7 +54,7 @@ Katanemo Arch-Function collection is built on top of the [Qwen 2.5](https://hugg
 
 
 ## Performance Benchmarks
-We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Leaderboard (BFCL)](https://gorilla.cs.berkeley.edu/leaderboard.html#leaderboard). For each model family, we select the one with the highest rank. The results (as of Oct 21st, 2024) are shwon below:
+We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Leaderboard (BFCL)](https://gorilla.cs.berkeley.edu/leaderboard.html#leaderboard). We compare with commonly-used models and the results (as of Oct 21st, 2024) are shwon below. For each model family, we select the one with the highest rank.
 
 <table>
   <tr style="text-align: center; vertical-align: middle; font-weight: bold;">
@@ -84,27 +84,16 @@ We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Lead
     <td>63.41%</td>
     <td>82.93%</td>
   </tr>
-  <tr style="text-align: center; vertical-align: middle;">
-    <td>2</td>
-    <td>Functionary-Medium-v3.1 (FC)</td>
-    <td>62.02%</td>
-    <td>89.52%</td>
-    <td>89.77%</td>
-    <td>73.48%</td>
-    <td>23.50%</td>
-    <td>70.73%</td>
-    <td>73.32%</td>
-  </tr>
-  <tr style="text-align: center; vertical-align: middle;">
-    <td>5</td>
-    <td>ToolACE-8B (FC)</td>
-    <td>60.44%</td>
-    <td>87.06%</td>
-    <td>89.52%</td>
-    <td>74.99%</td>
-    <td>17.38%</td>
-    <td>80.49%</td>
-    <td>85.71%</td>
+  <tr style="text-align: center; vertical-align: middle; font-weight: bold;">
+    <td> </td>
+    <td>Arch-Function-7B</td>
+    <td>59.62%</td>
+    <td>86.83%</td>
+    <td>88.07%</td>
+    <td>71.57%</td>
+    <td>21.00%</td>
+    <td>95.12%</td>
+    <td>73.63%</td>
   </tr>
   <tr style="text-align: center; vertical-align: middle;">
     <td>6</td>
@@ -117,28 +106,6 @@ We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Lead
     <td>73.17%</td>
     <td>74.60%</td>
   </tr>
-  <tr style="text-align: center; vertical-align: middle; font-weight: bold;">
-    <td> </td>
-    <td>Arch-Function-7B</td>
-    <td>58.44%</td>
-    <td>85.58%</td>
-    <td>88.14%</td>
-    <td>69.08%</td>
-    <td>20.50%</td>
-    <td>92.68%</td>
-    <td>74.05%</td>
-  </tr>
-  <tr style="text-align: center; vertical-align: middle; ">
-    <td>8</td>
-    <td>xLAM-8x22b-r (FC)</td>
-    <td>57.99%</td>
-    <td>88.15%</td>
-    <td>90.11%</td>
-    <td>71.97%</td>
-    <td>14.50%</td>
-    <td>85.37%</td>
-    <td>67.29%</td>
-  </tr>
   <tr style="text-align: center; vertical-align: middle; ">
     <td>9</td>
     <td>Gemini-1.5-Flash-002 (Prompt)</td>
@@ -150,16 +117,16 @@ We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Lead
     <td>85.37%</td>
     <td>78.54%</td>
   </tr>
-  <tr style="text-align: center; vertical-align: middle; ">
-    <td>10</td>
-    <td>Hammer2.0-7b (FC)</td>
+  <tr style="text-align: center; vertical-align: middle; font-weight: bold;">
+    <td> </td>
+    <td>Arch-Function-3B</td>
     <td>57.69%</td>
-    <td>90.27%</td>
-    <td>89.25%</td>
-    <td>69.79%</td>
-    <td>14.75%</td>
-    <td>95.12%</td>
-    <td>68.46%</td>
+    <td>85.19%</td>
+    <td>86.18%</td>
+    <td>71.21%</td>
+    <td>17.50%</td>
+    <td>90.24%</td>
+    <td>72.88%</td>
   </tr>
   <tr style="text-align: center; vertical-align: middle; ">
     <td>12</td>
@@ -183,50 +150,16 @@ We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Lead
     <td>75.61%</td>
     <td>49.44%</td>
   </tr>
-  <tr style="text-align: center; vertical-align: middle; font-weight: bold;">
-    <td> </td>
-    <td>Arch-Function-3B</td>
-    <td>56.57%</td>
-    <td>83.62%</td>
-    <td>85.36%</td>
-    <td>66.90%</td>
-    <td>19.50%</td>
-    <td>97.56%</td>
-    <td>70.99%</td>
-  </tr>
-  </tr>
   <tr style="text-align: center; vertical-align: middle; font-weight: bold;">
     <td> </td>
     <td>Arch-Function-1.5B</td>
-    <td>54.52%</td>
-    <td>80.31%</td>
-    <td>82.04%</td>
-    <td>66.19%</td>
-    <td>17.25%</td>
-    <td>97.56%</td>
-    <td>69.95%</td>
-  </tr>
-  <tr style="text-align: center; vertical-align: middle; ">
-    <td>19</td>
-    <td>xLAM-7b-r (FC)</td>
-    <td>54.41%</td>
-    <td>81.40%</td>
-    <td>83.46%</td>
-    <td>67.88%</td>
-    <td>14.50%</td>
-    <td>97.56%</td>
-    <td>64.05%</td>
-  </tr>
-  <tr style="text-align: center; vertical-align: middle; ">
-    <td>20</td>
-    <td>Qwen2.5-7B-Instruct (Prompt)</td>
-    <td>54.27%</td>
-    <td>85.79%</td>
-    <td>88.13%</td>
-    <td>65.97%</td>
-    <td>11.25%</td>
-    <td>92.68%</td>
-    <td>64.95%</td>
+    <td>56.20%</td>
+    <td>84.40%</td>
+    <td>83.96%</td>
+    <td>69.36%</td>
+    <td>15.88%</td>
+    <td>87.80%</td>
+    <td>74.39%</td>
   </tr>
   <tr style="text-align: center; vertical-align: middle; ">
     <td>21</td>
@@ -254,7 +187,7 @@ We evaluate Katanemo Arch-Function series on the [Berkeley Function-Calling Lead
 
 
 # Requirements
-The code of Arch-Function-1.5B has been in the Hugging Face `transformers` library and we advise you to install latest version:
+The code of Arch-Function-3B has been in the Hugging Face `transformers` library and we advise you to install latest version:
 ```bash
 pip install transformers>=4.37.0
 ```
@@ -270,7 +203,7 @@ import json
 from typing import Any, Dict, List
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
-model_name = "katanemo/Arch-Function-1.5B"
+model_name = "katanemo/Arch-Function-3B"
 model = AutoModelForCausalLM.from_pretrained(
     model_name, device_map="auto", torch_dtype="auto", trust_remote_code=True
 )
@@ -409,4 +342,4 @@ The current temperature in Seattle is 62 degrees in Fahrenheit.
 
 
 # License
-Katanemo Arch-Function collection is distributed under the [Katanemo license](https://huggingface.co/katanemolabs/Arch-Function-1.5B/blob/main/LICENSE).
+Katanemo Arch-Function collection is distributed under the [Katanemo license](https://huggingface.co/katanemolabs/Arch-Function-3B.gguf/blob/main/LICENSE).