End of training

README.md

@@ -20,7 +20,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [kiranpantha/whisper-large-v3-nepali](https://huggingface.co/kiranpantha/whisper-large-v3-nepali) on the OpenSLR54 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2635
+- Loss: 0.2699
 
 ## Model description
 
@@ -53,9 +53,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| No log        | 1.0   | 9    | 0.5598          |
-| No log        | 2.0   | 18   | 0.2679          |
-| 0.4551        | 3.0   | 27   | 0.2635          |
+| No log        | 1.0   | 14   | 0.2825          |
+| 0.5631        | 2.0   | 28   | 0.2673          |
+| 0.5631        | 3.0   | 42   | 0.2699          |
 
 
 ### Framework versions

adapter_config.json

@@ -22,11 +22,12 @@
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 128,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
     "v_proj",
+    "c_attn",
     "q_proj"
   ],
   "task_type": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61aa8a9f2b165ad065c069e7a07892c3d8e6bdf320a42567b3e2dba008b94614
-size 251714264
+oid sha256:fa0cceba3b25a13f70a571b7fb94e4757705e8789f0bb35617d9719c97db8822
+size 62969640

runs/Jan14_16-30-59_idc-training-gpu-compute-28/events.out.tfevents.1736872260.idc-training-gpu-compute-28.2914602.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80a0cd08578943c77daadfbec75917355d0e46af31abad66a1e210176ffd4c94
+size 7331

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6dcd769ea39d7a266b875de79854c7d5d8cfa6482d5ad4937d544f23bc9d4e85
+oid sha256:52476b8054287425867117be49a83b13225012c64a18ffdb0b6e7674a7df1fa8
 size 5560