asimokby/nllb-finetuned-de-en

README.md

@@ -5,18 +5,18 @@ base_model: facebook/nllb-200-distilled-600M
 tags:
 - generated_from_trainer
 model-index:
-- name: german
+- name: nllb-finetuned-de-en
   results: []
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# german
+# nllb-finetuned-de-en
 
-This model is a fine-tuned version of [facebook/nllb-200-distilled-600M](https://huggingface.co/facebook/nllb-200-distilled-600M) on the on the [TEDx dataset](https://huggingface.co/datasets/IWSLT/ted_talks_iwslt).
+This model is a fine-tuned version of [facebook/nllb-200-distilled-600M](https://huggingface.co/facebook/nllb-200-distilled-600M) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6058
+- Loss: 0.5475
 
 ## Model description
 
@@ -41,27 +41,31 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 10
+- num_epochs: 15
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.1885        | 1.0   | 503  | 0.9689          |
-| 0.9215        | 2.0   | 1006 | 0.8678          |
-| 0.8009        | 3.0   | 1509 | 0.7949          |
-| 0.7114        | 4.0   | 2012 | 0.7384          |
-| 0.6442        | 5.0   | 2515 | 0.6946          |
-| 0.5896        | 6.0   | 3018 | 0.6614          |
-| 0.5534        | 7.0   | 3521 | 0.6380          |
-| 0.5216        | 8.0   | 4024 | 0.6196          |
-| 0.5005        | 9.0   | 4527 | 0.6091          |
-| 0.4914        | 10.0  | 5030 | 0.6058          |
+| No log        | 1.0   | 486  | 1.1181          |
+| 1.3516        | 2.0   | 972  | 0.9976          |
+| 1.0459        | 3.0   | 1458 | 0.9073          |
+| 0.9073        | 4.0   | 1944 | 0.8393          |
+| 0.7901        | 5.0   | 2430 | 0.7772          |
+| 0.7072        | 6.0   | 2916 | 0.7292          |
+| 0.6262        | 7.0   | 3402 | 0.6872          |
+| 0.5713        | 8.0   | 3888 | 0.6532          |
+| 0.5185        | 9.0   | 4374 | 0.6228          |
+| 0.48          | 10.0  | 4860 | 0.5997          |
+| 0.4424        | 11.0  | 5346 | 0.5795          |
+| 0.4242        | 12.0  | 5832 | 0.5646          |
+| 0.4049        | 13.0  | 6318 | 0.5558          |
+| 0.3822        | 14.0  | 6804 | 0.5492          |
+| 0.3792        | 15.0  | 7290 | 0.5475          |
 
 
 ### Framework versions
 
 - Transformers 4.44.2
 - Pytorch 2.5.0+cu121
-- Datasets 3.1.0
 - Tokenizers 0.19.1

generation_config.json

@@ -3,7 +3,7 @@
   "bos_token_id": 0,
   "decoder_start_token_id": 2,
   "eos_token_id": 2,
-  "max_length": 256,
+  "max_length": 200,
   "pad_token_id": 1,
   "transformers_version": "4.44.2"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a762d1f5e0e9942a6fed214df42d992c238a778e40f248c1f39ece102ac91cb2
+oid sha256:936fbefecea41f74c69ba5169ec9f47b6547fc9689d3f5dd3e1482253d0c42dd
 size 2460354912

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f654c0c3f1c29080eb57a95498a1a057870727df51a0ee3efd7b1da5c61f774
-size 17331359
+oid sha256:7c0c8fca562d89c5dd1fb89f61b88a6208eeb89d4f5d1223ec54a2e827631bc0
+size 17331261

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e118795a46bf271e60de2b6cc574a46ae76d93d8870a1ef8382fbcd14791550b
-size 5368
+oid sha256:25a0608b6e32e88792b36881729a76ce1e12d0e7ff42072dd4e7482237a332ba
+size 5432