use none string not null

huggingface · Jan 8, 2025 · 1082595 · 1082595
1 parent c45c240
commit 1082595
Show file tree

Hide file tree

Showing 12 changed files with 46 additions and 57 deletions.
diff --git a/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb b/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb
@@ -60,7 +60,7 @@
     "import os\n",
     "from transformers import AutoModelForCausalLM, AutoTokenizer\n",
     "from datasets import load_dataset\n",
-    "from trl import DPOTrainer, DPOConfig\n"
+    "from trl import DPOTrainer, DPOConfig"
    ]
   },
   {
@@ -122,9 +122,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Model to fine-tune\n",
@@ -269,7 +267,7 @@
     "    # Use bfloat16 precision for faster training\n",
     "    bf16=True,\n",
     "    # Disable wandb/tensorboard logging\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Keep all columns in dataset even if not used\n",
     "    remove_unused_columns=False,\n",
     "    # Enable MPS (Metal Performance Shaders) for Mac devices\n",
@@ -282,7 +280,7 @@
     "    # Maximum length of the input prompt in tokens\n",
     "    max_prompt_length=1024,\n",
     "    # Maximum combined length of prompt + response in tokens\n",
-    "    max_length=1536\n",
+    "    max_length=1536,\n",
     ")"
    ]
   },
@@ -303,11 +301,11 @@
     "    processing_class=tokenizer,\n",
     "    # DPO-specific temperature parameter that controls the strength of the preference model\n",
     "    # Lower values (like 0.1) make the model more conservative in following preferences\n",
-    "    #beta=0.1,\n",
+    "    # beta=0.1,\n",
     "    # Maximum length of the input prompt in tokens\n",
-    "    #max_prompt_length=1024,\n",
+    "    # max_prompt_length=1024,\n",
     "    # Maximum combined length of prompt + response in tokens\n",
-    "    #max_length=1536,\n",
+    "    # max_length=1536,\n",
     ")"
    ]
   },
@@ -364,7 +362,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.12.7"
+   "version": "3.11.10"
   },
   "widgets": {
    "application/vnd.jupyter.widget-state+json": {

diff --git a/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb b/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb
@@ -348,7 +348,7 @@
     "    # Gradual learning rate warmup\n",
     "    warmup_steps=10,\n",
     "    # Disable external logging\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Where to save model/checkpoints\n",
     "    output_dir=\"./results/\",\n",
     "    # Enable MPS (Metal Performance Shaders) if available\n",

diff --git a/3_parameter_efficient_finetuning/notebooks/finetune_sft_peft.ipynb b/3_parameter_efficient_finetuning/notebooks/finetune_sft_peft.ipynb
@@ -249,7 +249,7 @@
     "    bf16=True,  # Use bfloat16 precision\n",
     "    # Integration settings\n",
     "    push_to_hub=False,  # Don't push to HuggingFace Hub\n",
-    "    report_to=None,  # Disable external logging\n",
+    "    report_to=\"none\",  # Disable external logging\n",
     ")"
    ]
   },

diff --git a/es/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb b/es/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb
@@ -34,7 +34,7 @@
     "\n",
     "login()\n",
     "\n",
-    "# Para conveniencia, puedes crear una variable de entorno que contenga tu token de Hugging Face como HF_TOKEN\n"
+    "# Para conveniencia, puedes crear una variable de entorno que contenga tu token de Hugging Face como HF_TOKEN"
    ]
   },
   {
@@ -60,7 +60,7 @@
     "import os\n",
     "from transformers import AutoModelForCausalLM, AutoTokenizer\n",
     "from datasets import load_dataset\n",
-    "from trl import DPOTrainer, DPOConfig\n"
+    "from trl import DPOTrainer, DPOConfig"
    ]
   },
   {
@@ -121,9 +121,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Modelo para afinar\n",
@@ -137,7 +135,7 @@
     "\n",
     "# Establece el nombre para el modelo afinado que se guardará y/o subirá\n",
     "finetune_name = \"SmolLM2-FT-DPO\"\n",
-    "finetune_tags = [\"smol-course\", \"module_2\"]\n"
+    "finetune_tags = [\"smol-course\", \"module_2\"]"
    ]
   },
   {
@@ -268,14 +266,14 @@
     "    # Usar precisión bfloat16 para un entrenamiento más rápido\n",
     "    bf16=True,\n",
     "    # Desactivar los registros de wandb/tensorboard\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Mantener todas las columnas en el dataset incluso si no se utilizan\n",
     "    remove_unused_columns=False,\n",
     "    # Habilitar MPS (Metal Performance Shaders) para dispositivos Mac\n",
     "    use_mps_device=device == \"mps\",\n",
     "    # ID del modelo para cargas al HuggingFace Hub\n",
     "    hub_model_id=finetune_name,\n",
-    ")\n"
+    ")"
    ]
   },
   {
@@ -300,7 +298,7 @@
     "    max_prompt_length=1024,\n",
     "    # Longitud máxima combinada de prompt + respuesta en tokens\n",
     "    max_length=1536,\n",
-    ")\n"
+    ")"
    ]
   },
   {
@@ -317,7 +315,7 @@
     "\n",
     "# Sube el modelo a Hugging Face si el inicio de sesión o HF_TOKEN está configurado\n",
     "if os.getenv(\"HF_TOKEN\"):\n",
-    "    trainer.push_to_hub(tags=finetune_tags)\n"
+    "    trainer.push_to_hub(tags=finetune_tags)"
    ]
   },
   {

diff --git a/es/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb b/es/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb
@@ -59,6 +59,7 @@
     "\n",
     "# Autenticación en Hugging Face\n",
     "from huggingface_hub import login\n",
+    "\n",
     "login()"
    ]
   },
@@ -258,9 +259,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Modelo para afinar\n",
@@ -274,7 +273,7 @@
     "\n",
     "# Establece el nombre para guardar o subir el modelo afinado\n",
     "finetune_name = \"SmolLM2-FT-DPO\"\n",
-    "finetune_tags = [\"smol-course\", \"module_2\"]\n"
+    "finetune_tags = [\"smol-course\", \"module_2\"]"
    ]
   },
   {
@@ -350,13 +349,13 @@
     "    # Calentamiento gradual de la tasa de aprendizaje\n",
     "    warmup_steps=10,\n",
     "    # Desactivar registro externo\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Dónde guardar el modelo/puntos de control\n",
     "    output_dir=\"./results/\",\n",
     "    # Habilitar MPS (Metal Performance Shaders) si está disponible\n",
     "    use_mps_device=device == \"mps\",\n",
     "    hub_model_id=finetune_name,\n",
-    ")\n"
+    ")"
    ]
   },
   {

diff --git a/es/3_parameter_efficient_finetuning/notebooks/finetune_sft_peft.ipynb b/es/3_parameter_efficient_finetuning/notebooks/finetune_sft_peft.ipynb
@@ -49,7 +49,7 @@
     "\n",
     "login()\n",
     "\n",
-    "# Por comodidad, puedes crear una variable de entorno que contenga tu token de Hugging Face como HF_TOKEN\n"
+    "# Por comodidad, puedes crear una variable de entorno que contenga tu token de Hugging Face como HF_TOKEN"
    ]
   },
   {
@@ -249,7 +249,7 @@
     "    bf16=True,  # Usar precisión bfloat16\n",
     "    # Configuración de integración\n",
     "    push_to_hub=False,  # No subir a HuggingFace\n",
-    "    report_to=None,  # Desactivar registro externo\n",
+    "    report_to=\"none\",  # Desactivar registro externo\n",
     ")"
    ]
   },

diff --git a/ko/3_parameter_efficient_finetuning/notebooks/finetune_sft_peft.ipynb b/ko/3_parameter_efficient_finetuning/notebooks/finetune_sft_peft.ipynb
@@ -250,7 +250,7 @@
     "    bf16=True,  # bfloat16 정밀도 사용\n",
     "    # 통합 설정\n",
     "    push_to_hub=False,  # Hugging Face 허브로 내보내지 않음\n",
-    "    report_to=None,  # 외부 로깅 비활성화\n",
+    "    report_to=\"none\",  # 외부 로깅 비활성화\n",
     ")"
    ]
   },
@@ -284,7 +284,7 @@
     "    packing=True,  # 효율성을 위해 입력 패킹 활성화\n",
     "    dataset_kwargs={\n",
     "        \"add_special_tokens\": False,  # 템플릿에서 추가 토큰 처리\n",
-    "        \"append_concat_token\": False,  # 추가 구분자 없음 \n",
+    "        \"append_concat_token\": False,  # 추가 구분자 없음\n",
     "    },\n",
     ")"
    ]

diff --git a/pt-br/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb b/pt-br/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb
@@ -60,7 +60,7 @@
     "import os\n",
     "from transformers import AutoModelForCausalLM, AutoTokenizer\n",
     "from datasets import load_dataset\n",
-    "from trl import DPOTrainer, DPOConfig\n"
+    "from trl import DPOTrainer, DPOConfig"
    ]
   },
   {
@@ -122,9 +122,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Model to fine-tune\n",
@@ -269,7 +267,7 @@
     "    # Use bfloat16 precision for faster training\n",
     "    bf16=True,\n",
     "    # Disable wandb/tensorboard logging\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Keep all columns in dataset even if not used\n",
     "    remove_unused_columns=False,\n",
     "    # Enable MPS (Metal Performance Shaders) for Mac devices\n",

diff --git a/pt-br/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb b/pt-br/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb
@@ -58,6 +58,7 @@
     "\n",
     "# Authenticate to Hugging Face\n",
     "from huggingface_hub import login\n",
+    "\n",
     "login()"
    ]
   },
@@ -257,9 +258,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Model to fine-tune\n",
@@ -349,7 +348,7 @@
     "    # Gradual learning rate warmup\n",
     "    warmup_steps=10,\n",
     "    # Disable external logging\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Where to save model/checkpoints\n",
     "    output_dir=\"./results/\",\n",
     "    # Enable MPS (Metal Performance Shaders) if available\n",

diff --git a/vi/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb b/vi/2_preference_alignment/notebooks/dpo_finetuning_example.ipynb
@@ -117,9 +117,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Mô hình để tinh chỉnh\n",
@@ -264,7 +262,7 @@
     "    # Sử dụng bfloat16 để tăng tốc huấn luyện\n",
     "    bf16=True,\n",
     "    # Tắt logging wandb/tensorboard\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Giữ lại tất cả các cột trong dataset kể cả không sử dụng\n",
     "    remove_unused_columns=False,\n",
     "    # Bật MPS cho thiết bị Mac\n",

diff --git a/vi/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb b/vi/2_preference_alignment/notebooks/orpo_finetuning_example.ipynb
@@ -53,6 +53,7 @@
     "\n",
     "# Authenticate to Hugging Face\n",
     "from huggingface_hub import login\n",
+    "\n",
     "login()"
    ]
   },
@@ -90,7 +91,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# TODO: 🐕 Nếu bộ dữ liệu của bạn không được biểu diễn dưới dạng danh sách hội thoại, \n",
+    "# TODO: 🐕 Nếu bộ dữ liệu của bạn không được biểu diễn dưới dạng danh sách hội thoại,\n",
     "# bạn có thể sử dụng hàm `process_dataset` để chuyển đổi nó."
    ]
   },
@@ -248,9 +249,7 @@
     "device = (\n",
     "    \"cuda\"\n",
     "    if torch.cuda.is_available()\n",
-    "    else \"mps\"\n",
-    "    if torch.backends.mps.is_available()\n",
-    "    else \"cpu\"\n",
+    "    else \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
     ")\n",
     "\n",
     "# Mô hình để tinh chỉnh\n",
@@ -315,7 +314,7 @@
     "    # Learning rate nhỏ để tránh catastrophic forgetting\n",
     "    # Đọc thêm về catastrophic forgetting: https://blog.vinbigdata.org/robot-co-the-hoc-suot-doi-khong/\n",
     "    learning_rate=8e-6,\n",
-    "    # Giảm learning rate tuyến tính trong quá trình huấn luyện  \n",
+    "    # Giảm learning rate tuyến tính trong quá trình huấn luyện\n",
     "    lr_scheduler_type=\"linear\",\n",
     "    # Độ dài tối đa kết hợp của chỉ thị + câu trả lời\n",
     "    max_length=1024,\n",
@@ -333,15 +332,15 @@
     "    # Số epoch huấn luyện\n",
     "    num_train_epochs=1,\n",
     "    # Khi nào chạy đánh giá\n",
-    "    evaluation_strategy=\"steps\", \n",
+    "    evaluation_strategy=\"steps\",\n",
     "    # Đánh giá sau mỗi 20% huấn luyện\n",
     "    eval_steps=0.2,\n",
     "    # Ghi log metrics mỗi step\n",
     "    logging_steps=1,\n",
     "    # Tăng learning rate dần dần\n",
     "    warmup_steps=10,\n",
     "    # Tắt logging bên ngoài\n",
-    "    report_to=None,\n",
+    "    report_to=\"none\",\n",
     "    # Nơi lưu model/checkpoints\n",
     "    output_dir=\"./results/\",\n",
     "    # Bật MPS (Metal Performance Shaders) nếu có\n",