ninagroot/Llama-360Mtest

README.md

@@ -13,7 +13,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.5128
+- Loss: 3.6292
 
 ## Model description
 
@@ -41,15 +41,18 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_steps: 100
-- num_epochs: 2
+- num_epochs: 5
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 2.9191        | 1.0   | 254  | 3.7445          |
-| 1.828         | 2.0   | 508  | 3.5128          |
+| 2.9655        | 1.0   | 254  | 3.7779          |
+| 1.7081        | 2.0   | 509  | 3.5625          |
+| 0.8711        | 3.0   | 763  | 3.5947          |
+| 0.4644        | 4.0   | 1018 | 3.6139          |
+| 0.273         | 4.99  | 1270 | 3.6292          |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06c31e469dafc14f5b44973fff506cb6b58c7e9a29212eca1a94b10698acd27e
+oid sha256:08ff4a71f92b2cd5ce55ee89dd8685f14b0fa5fc2140efa0a6db6536c1e2bf79
 size 1570992472

runs/Apr04_11-16-01_gcn68.local.snellius.surf.nl/events.out.tfevents.1712222170.gcn68.local.snellius.surf.nl.1796157.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7065e7e77daffd672959e3cfcfdf16d7d49b9b30398d759d826ea78cdf9910c0
+size 19540

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:56176203a926a2239766a855e8a4c7e560543261d6afbdedcc1dfba1e0c75197
+oid sha256:0de1b160a42adb1fc414130b23235d6f2ddc421aca0bc43a8b43a392362b2944
 size 4984