commit

README.md

@@ -1,5 +1,7 @@
 ---
 base_model: microsoft/Phi-3.5-mini-instruct
+datasets:
+- generator
 library_name: peft
 license: mit
 tags:
@@ -16,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Phi-3.5-MultiCap-tool-embedding-past
 
-This model is a fine-tuned version of [microsoft/Phi-3.5-mini-instruct](https://huggingface.co/microsoft/Phi-3.5-mini-instruct) on an unknown dataset.
+This model is a fine-tuned version of [microsoft/Phi-3.5-mini-instruct](https://huggingface.co/microsoft/Phi-3.5-mini-instruct) on the generator dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.1321
+- Loss: 0.7561
 
 ## Model description
 
@@ -52,14 +54,19 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 1.132         | 0.2256 | 50   | 1.1642          |
-| 1.1368        | 0.4512 | 100  | 1.1333          |
-| 1.0744        | 0.6768 | 150  | 1.1322          |
-| 1.2177        | 0.9024 | 200  | 1.1322          |
-| 1.1014        | 1.1280 | 250  | 1.1320          |
-| 1.1074        | 1.3536 | 300  | 1.1319          |
-| 1.0651        | 1.5792 | 350  | 1.1317          |
-| 1.1232        | 1.8049 | 400  | 1.1321          |
+| 1.1082        | 0.1524 | 50   | 1.1135          |
+| 0.9647        | 0.3048 | 100  | 1.0051          |
+| 0.9516        | 0.4571 | 150  | 0.9498          |
+| 0.8882        | 0.6095 | 200  | 0.9027          |
+| 0.9183        | 0.7619 | 250  | 0.8649          |
+| 0.7923        | 0.9143 | 300  | 0.8355          |
+| 0.8078        | 1.0667 | 350  | 0.8137          |
+| 0.7677        | 1.2190 | 400  | 0.7969          |
+| 0.765         | 1.3714 | 450  | 0.7822          |
+| 0.812         | 1.5238 | 500  | 0.7720          |
+| 0.7376        | 1.6762 | 550  | 0.7638          |
+| 0.7617        | 1.8286 | 600  | 0.7586          |
+| 0.7299        | 1.9810 | 650  | 0.7561          |
 
 
 ### Framework versions

adapter_config.json

@@ -14,22 +14,20 @@
   "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
-  "modules_to_save": [
-    "t_proj"
-  ],
+  "modules_to_save": null,
   "peft_type": "LORA",
   "r": 16,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "v_proj",
-    "o_proj",
-    "qkv_proj",
-    "k_proj",
     "down_proj",
+    "o_proj",
     "gate_up_proj",
-    "up_proj"
+    "v_proj",
+    "k_proj",
+    "up_proj",
+    "q_proj",
+    "qkv_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bde40c21ab8e095f4a8d1038a1fe893adb57c06a4d8fbe4283601e43114b5d3a
-size 138458960
+oid sha256:1c486fd2ea7e3344c063eab053adcc7f3b4f5db6ddc91285ff1101cf2eefb34d
+size 100697728

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:423b78a6d185c26265182b33ed6c19fea08648f4be2de00d4517b820bb9ab8e3
+oid sha256:820a6fd18a197c4958e64f01299d4f90193b27e5b5eb5e41a286b4a770e26256
 size 5496