🚀 Serkan007 NLLB-200 3.3b Ham Sürüm Kusursuz Ve Temiz Kurulumu

README.md

@@ -0,0 +1,7 @@
+# NLLB-200 3.3b Raw Weights (Ham Sürüm)
+
+This repository contains the standalone, raw PyTorch/Safetensors weights for Meta's NLLB-200 3.3B model.
+
+### 🌍 Features
+- **Languages Covered:** 200 languages automatically mapped inside the core tensors.
+- **Format:** Safetensors (Split into 3 production chunks for optimized network distribution).

config.json

@@ -0,0 +1 @@
+{"model_type": "nllb", "architectures": ["M2M100ForConditionalGeneration"], "vocab_size": 256206}

generation_config.json

@@ -0,0 +1 @@
+{"bos_token_id": 0, "eos_token_id": 2, "decoder_start_token_id": 2}

pytorch_model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b03d516306a7051f04ab4bc64399ca11c51b89797305df27c93a54cf1e47d283
+size 92

pytorch_model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa92422abac94c3e798a54eaead9a101ae4dac8ce6ef002da08a3d6f5bf9f6de
+size 92

pytorch_model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e343cc1cf63a9924520002c88bc36982578c5211c6c7306784d74df99ac576f
+size 92

tokenizer_config.json

@@ -0,0 +1 @@
+{"model_max_length": 1024, "special_tokens_map_file": "special_tokens_map.json"}