Added fineTuned Model of Helsinki

app.py

@@ -9,6 +9,7 @@ from transformers.pipelines.audio_utils import ffmpeg_read
 
 import tempfile
 import os
+import safetensors
 
 MODEL_NAME = "openai/whisper-large-v3-turbo"
 BATCH_SIZE = 8
@@ -25,8 +26,15 @@ pipe = pipeline(
 )
 
 model_name_translate = "Helsinki-NLP/opus-mt-en-ar"
+model_path_translate = "./lora-opus-mt-en-ar"
+tensors = {}
+
+with safe_open("lora-opus-mt-en-ar\model.safetensors", framework="pt", device=0) as f:
+    for k in f.keys():
+        tensors[k] = f.get_tensor(k)
+        
 tokenizer_translation = MarianTokenizer.from_pretrained(model_name_translate)
-model_translate = MarianMTModel.from_pretrained(model_name_translate)
+model_translate = MarianMTModel.from_pretrained(model_path_translate, state_dict=tensors, config="lora-opus-mt-en-ar/config.json")
 
 @spaces.GPU
 def translate(sentence):

lora-opus-mt-en-ar/adapter_config.json

@@ -0,0 +1,8 @@
+{
+  "r": 8,
+  "lora_alpha": 4,
+  "target_modules": ["q_proj", "v_proj"],
+  "lora_dropout": 0.1,
+  "bias": "none"
+
+}

lora-opus-mt-en-ar/config.json

@@ -0,0 +1,61 @@
+{
+  "_name_or_path": "./lora-opus-mt-en-ar",
+  "activation_dropout": 0.0,
+  "activation_function": "swish",
+  "add_bias_logits": false,
+  "add_final_layer_norm": false,
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "bad_words_ids": [
+    [
+      62801
+    ]
+  ],
+  "bos_token_id": 0,
+  "classif_dropout": 0.0,
+  "classifier_dropout": 0.0,
+  "d_model": 512,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 62801,
+  "decoder_vocab_size": 62802,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 0,
+  "extra_pos_embeddings": 62802,
+  "forced_eos_token_id": 0,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2"
+  },
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1,
+    "LABEL_2": 2
+  },
+  "max_length": 512,
+  "max_position_embeddings": 512,
+  "model_type": "marian",
+  "normalize_before": false,
+  "normalize_embedding": false,
+  "num_beams": 4,
+  "num_hidden_layers": 6,
+  "pad_token_id": 62801,
+  "scale_embedding": true,
+  "share_encoder_decoder_embeddings": true,
+  "static_position_embeddings": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.2",
+  "use_cache": true,
+  "vocab_size": 62802
+}

lora-opus-mt-en-ar/generation_config.json

@@ -0,0 +1,16 @@
+{
+  "bad_words_ids": [
+    [
+      62801
+    ]
+  ],
+  "bos_token_id": 0,
+  "decoder_start_token_id": 62801,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "max_length": 512,
+  "num_beams": 4,
+  "pad_token_id": 62801,
+  "renormalize_logits": true,
+  "transformers_version": "4.44.2"
+}

lora-opus-mt-en-ar/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a26674db967a059bb43099392160471d21739f3372b5a7a5b7a624c0c62261f
+size 305452744

lora-opus-mt-en-ar/special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

lora-opus-mt-en-ar/tokenizer_config.json

@@ -0,0 +1,38 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "62801": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "separate_vocabs": false,
+  "source_lang": "eng",
+  "sp_model_kwargs": {},
+  "target_lang": "ara",
+  "tokenizer_class": "MarianTokenizer",
+  "unk_token": "<unk>"
+}

requirements.txt

@@ -3,3 +3,4 @@ yt-dlp
 torch
 sentencePiece
 sacremoses
+safetensors