Upload 4 files

Alpaca_+_Llama_7b_full_example.ipynb

@@ -35,9 +35,7 @@
         "else:\n",
         "    # Use this for older GPUs (V100, Tesla T4, RTX 20xx)\n",
         "    !pip install \"unsloth[colab] @ git+https://github.com/unslothai/unsloth.git\"\n",
-        "pass\n",
-        "\n",
-        "!pip install \"git+https://github.com/huggingface/transformers.git\" # Native 4bit loading works!"
+        "pass"
       ]
     },
     {
@@ -280,11 +278,12 @@
         "# 4bit pre quantized models we support for 4x faster downloading + no OOMs.\n",
         "fourbit_models = [\n",
         "    \"unsloth/mistral-7b-bnb-4bit\",\n",
+        "    \"unsloth/mistral-7b-instruct-v0.2-bnb-4bit\",\n",
         "    \"unsloth/llama-2-7b-bnb-4bit\",\n",
         "    \"unsloth/llama-2-13b-bnb-4bit\",\n",
         "    \"unsloth/codellama-34b-bnb-4bit\",\n",
         "    \"unsloth/tinyllama-bnb-4bit\",\n",
-        "]\n",
+        "] # More models at https://huggingface.co/unsloth\n",
         "\n",
         "model, tokenizer = FastLanguageModel.from_pretrained(\n",
         "    model_name = \"unsloth/llama-2-7b-bnb-4bit\", # Choose ANY! eg mistralai/Mistral-7B-Instruct-v0.2\n",
@@ -348,7 +347,11 @@
         "\n",
         "**[NOTE]** To train only on completions (ignoring the user's input) read TRL's docs [here](https://huggingface.co/docs/trl/sft_trainer#train-on-completions-only).\n",
         "\n",
-        "**[NOTE]** Remember to add the **EOS_TOKEN** to the tokenized output!! Otherwise you'll get infinite generations!"
+        "**[NOTE]** Remember to add the **EOS_TOKEN** to the tokenized output!! Otherwise you'll get infinite generations!\n",
+        "\n",
+        "If you want to use the `ChatML` template for ShareGPT datasets, try our conversational [notebook](https://colab.research.google.com/drive/1Aau3lgPzeZKQ-98h69CCu1UJcvIBLmy2?usp=sharing).\n",
+        "\n",
+        "For text completions like novel writing, try this [notebook](https://colab.research.google.com/drive/1ef-tab5bhkvWmBOObepl1WgJvfvSzn5Q?usp=sharing)."
       ],
       "metadata": {
         "id": "vITh0KVJ10qX"
@@ -656,6 +659,7 @@
         "\n",
         "trainer = SFTTrainer(\n",
         "    model = model,\n",
+        "    tokenizer = tokenizer,\n",
         "    train_dataset = dataset,\n",
         "    dataset_text_field = \"text\",\n",
         "    max_seq_length = max_seq_length,\n",
@@ -1054,7 +1058,7 @@
       "cell_type": "code",
       "source": [
         "# alpaca_prompt = Copied from above\n",
-        "\n",
+        "FastLanguageModel.for_inference(model) # Enable native 2x faster inference\n",
         "inputs = tokenizer(\n",
         "[\n",
         "    alpaca_prompt.format(\n",
@@ -1062,7 +1066,7 @@
         "        \"1, 1, 2, 3, 5, 8\", # input\n",
         "        \"\", # output - leave this blank for generation!\n",
         "    )\n",
-        "]*1, return_tensors = \"pt\").to(\"cuda\")\n",
+        "], return_tensors = \"pt\").to(\"cuda\")\n",
         "\n",
         "outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True)\n",
         "tokenizer.batch_decode(outputs)"
@@ -1101,7 +1105,7 @@
       "cell_type": "code",
       "source": [
         "# alpaca_prompt = Copied from above\n",
-        "\n",
+        "FastLanguageModel.for_inference(model) # Enable native 2x faster inference\n",
         "inputs = tokenizer(\n",
         "[\n",
         "    alpaca_prompt.format(\n",
@@ -1109,7 +1113,7 @@
         "        \"1, 1, 2, 3, 5, 8\", # input\n",
         "        \"\", # output - leave this blank for generation!\n",
         "    )\n",
-        "]*1, return_tensors = \"pt\").to(\"cuda\")\n",
+        "], return_tensors = \"pt\").to(\"cuda\")\n",
         "\n",
         "from transformers import TextStreamer\n",
         "text_streamer = TextStreamer(tokenizer)\n",
@@ -1187,6 +1191,7 @@
         "        dtype = dtype,\n",
         "        load_in_4bit = load_in_4bit,\n",
         "    )\n",
+        "    FastLanguageModel.for_inference(model) # Enable native 2x faster inference\n",
         "\n",
         "# alpaca_prompt = You MUST copy from above!\n",
         "\n",
@@ -1197,7 +1202,7 @@
         "        \"\", # input\n",
         "        \"\", # output - leave this blank for generation!\n",
         "    )\n",
-        "]*1, return_tensors = \"pt\").to(\"cuda\")\n",
+        "], return_tensors = \"pt\").to(\"cuda\")\n",
         "\n",
         "from transformers import TextStreamer\n",
         "text_streamer = TextStreamer(tokenizer)\n",
@@ -1227,7 +1232,7 @@
     {
       "cell_type": "markdown",
       "source": [
-        "You can also use Hugging Face's `AutoModelForPeftCausalLM`"
+        "You can also use Hugging Face's `AutoModelForPeftCausalLM`. Only use this if you do not have `unsloth` installed. It can be hopelessly slow, since `4bit` model downloading is not supported, and Unsloth's **inference is 2x faster**."
       ],
       "metadata": {
         "id": "TGKU509CuMmq"
@@ -1237,6 +1242,7 @@
       "cell_type": "code",
       "source": [
         "if False:\n",
+        "    # I highly do NOT suggest - use Unsloth if possible\n",
         "    from peft import AutoModelForPeftCausalLM\n",
         "    from transformers import AutoTokenizer\n",
         "    model = AutoModelForPeftCausalLM.from_pretrained(\n",
@@ -1256,7 +1262,7 @@
       "source": [
         "### Saving to float16 for VLLM\n",
         "\n",
-        "We also support saving to `float16` directly. Select `merged_16bit` for float16 or `merged_4bit` for int4. We also allow `lora` adapters as a fallback. Use `push_to_hub_merged` to upload to your Hugging Face account!"
+        "We also support saving to `float16` directly. Select `merged_16bit` for float16 or `merged_4bit` for int4. We also allow `lora` adapters as a fallback. Use `push_to_hub_merged` to upload to your Hugging Face account! You can go to https://huggingface.co/settings/tokens for your personal tokens."
       ],
       "metadata": {
         "id": "-xp0YDnKuN98"
@@ -1266,16 +1272,16 @@
       "cell_type": "code",
       "source": [
         "# Merge to 16bit\n",
-        "if False: model.save_pretrained_merged(\"x\", tokenizer, save_method = \"merged_16bit\",)\n",
-        "if False: model.push_to_hub_merged(\"hf_user/x\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
+        "if False: model.save_pretrained_merged(\"model\", tokenizer, save_method = \"merged_16bit\",)\n",
+        "if False: model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
         "\n",
         "# Merge to 4bit\n",
-        "if False: model.save_pretrained_merged(\"x\", tokenizer, save_method = \"merged_4bit\",)\n",
-        "if False: model.push_to_hub_merged(\"hf_user/x\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
+        "if False: model.save_pretrained_merged(\"model\", tokenizer, save_method = \"merged_4bit\",)\n",
+        "if False: model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
         "\n",
         "# Just LoRA adapters\n",
-        "if False: model.save_pretrained_merged(\"x\", tokenizer, save_method = \"lora\",)\n",
-        "if False: model.push_to_hub_merged(\"hf_user/x\", tokenizer, save_method = \"lora\", token = \"\")"
+        "if False: model.save_pretrained_merged(\"model\", tokenizer, save_method = \"lora\",)\n",
+        "if False: model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"lora\", token = \"\")"
       ],
       "metadata": {
         "id": "vnFt-4ymuPM1"
@@ -1287,7 +1293,12 @@
       "cell_type": "markdown",
       "source": [
         "### GGUF / llama.cpp Conversion\n",
-        "To save to `GGUF` / `llama.cpp`, we support it natively now! We clone `llama.cpp` and we default save it to `q8_0`. We allow all methods like `q4_k_m`. Use `save_pretrained_gguf` for local saving and `push_to_hub_gguf` for uploading to HF."
+        "To save to `GGUF` / `llama.cpp`, we support it natively now! We clone `llama.cpp` and we default save it to `q8_0`. We allow all methods like `q4_k_m`. Use `save_pretrained_gguf` for local saving and `push_to_hub_gguf` for uploading to HF.\n",
+        "\n",
+        "Some supported quant methods (full list on our [Wiki page](https://github.com/unslothai/unsloth/wiki#gguf-quantization-options)):\n",
+        "* `q8_0` - Fast conversion. High resource use, but generally acceptable.\n",
+        "* `q4_k_m` - Recommended. Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q4_K.\n",
+        "* `q5_k_m` - Recommended. Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q5_K."
       ],
       "metadata": {
         "id": "8xg8B-N7uQcE"
@@ -1297,16 +1308,16 @@
       "cell_type": "code",
       "source": [
         "# Save to 8bit Q8_0\n",
-        "if False: model.save_pretrained_gguf(\"x\", tokenizer,)\n",
-        "if False: model.push_to_hub_gguf(\"hf_user/x\", tokenizer, token = \"\")\n",
+        "if False: model.save_pretrained_gguf(\"model\", tokenizer,)\n",
+        "if False: model.push_to_hub_gguf(\"hf/model\", tokenizer, token = \"\")\n",
         "\n",
         "# Save to 16bit GGUF\n",
-        "if False: model.save_pretrained_gguf(\"x\", tokenizer, quantization_method = \"f16\")\n",
-        "if False: model.push_to_hub_gguf(\"hf_user/x\", tokenizer, quantization_method = \"f16\", token = \"\")\n",
+        "if False: model.save_pretrained_gguf(\"model\", tokenizer, quantization_method = \"f16\")\n",
+        "if False: model.push_to_hub_gguf(\"hf/model\", tokenizer, quantization_method = \"f16\", token = \"\")\n",
         "\n",
         "# Save to q4_k_m GGUF\n",
-        "if False: model.save_pretrained_gguf(\"x\", tokenizer, quantization_method = \"q4_k_m\")\n",
-        "if False: model.push_to_hub_gguf(\"hf_user/x\", tokenizer, quantization_method = \"q4_k_m\", token = \"\")"
+        "if False: model.save_pretrained_gguf(\"model\", tokenizer, quantization_method = \"q4_k_m\")\n",
+        "if False: model.push_to_hub_gguf(\"hf/model\", tokenizer, quantization_method = \"q4_k_m\", token = \"\")"
       ],
       "metadata": {
         "id": "8T822D9fuR0g"
@@ -1317,7 +1328,7 @@
     {
       "cell_type": "markdown",
       "source": [
-        "Now, use the `x.gguf` file or `x-unsloth-Q4_K_M.gguf` file in `llama.cpp` or a UI based system like `GPT4All`. You can install GPT4All by going [here](https://gpt4all.io/index.html)."
+        "Now, use the `model-unsloth.gguf` file or `model-unsloth-Q4_K_M.gguf` file in `llama.cpp` or a UI based system like `GPT4All`. You can install GPT4All by going [here](https://gpt4all.io/index.html)."
       ],
       "metadata": {
         "id": "RiRcv_rquUq0"
@@ -1333,8 +1344,10 @@
         "2. Mistral 7b 2x faster [free Colab](https://colab.research.google.com/drive/1Dyauq4kTZoLewQ1cApceUQVNcnnNTzg_?usp=sharing)\n",
         "3. TinyLlama 4x faster full Alpaca 52K in 1 hour [free Colab](https://colab.research.google.com/drive/1AZghoNBQaMDgWJpi4RbffGM1h6raLUj9?usp=sharing)\n",
         "4. CodeLlama 34b 2x faster [A100 on Colab](https://colab.research.google.com/drive/1y7A0AxE3y8gdj4AVkl2aZX47Xu3P1wJT?usp=sharing)\n",
-        "5. Llama 7b [free Kaggle](https://www.kaggle.com/danielhanchen/unsloth-alpaca-t4-ddp)\n",
+        "5. Mistral 7b [free Kaggle version](https://www.kaggle.com/code/danielhanchen/kaggle-mistral-7b-unsloth-notebook)\n",
         "6. We also did a [blog](https://huggingface.co/blog/unsloth-trl) with 🤗 HuggingFace, and we're in the TRL [docs](https://huggingface.co/docs/trl/main/en/sft_trainer#accelerate-fine-tuning-2x-using-unsloth)!\n",
+        "7. `ChatML` for ShareGPT datasets, [conversational notebook](https://colab.research.google.com/drive/1Aau3lgPzeZKQ-98h69CCu1UJcvIBLmy2?usp=sharing)\n",
+        "8. Text completions like novel writing [notebook](https://colab.research.google.com/drive/1ef-tab5bhkvWmBOObepl1WgJvfvSzn5Q?usp=sharing)\n",
         "\n",
         "<div class=\"align-center\">\n",
         "  <a href=\"https://github.com/unslothai/unsloth\"><img src=\"https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png\" width=\"115\"></a>\n",

Alpaca_+_Mistral_7b_full_example.ipynb

@@ -35,9 +35,7 @@
         "else:\n",
         "    # Use this for older GPUs (V100, Tesla T4, RTX 20xx)\n",
         "    !pip install \"unsloth[colab] @ git+https://github.com/unslothai/unsloth.git\"\n",
-        "pass\n",
-        "\n",
-        "!pip install \"git+https://github.com/huggingface/transformers.git\" # Native 4bit loading works!"
+        "pass"
       ]
     },
     {
@@ -280,14 +278,15 @@
         "# 4bit pre quantized models we support for 4x faster downloading + no OOMs.\n",
         "fourbit_models = [\n",
         "    \"unsloth/mistral-7b-bnb-4bit\",\n",
+        "    \"unsloth/mistral-7b-instruct-v0.2-bnb-4bit\",\n",
         "    \"unsloth/llama-2-7b-bnb-4bit\",\n",
         "    \"unsloth/llama-2-13b-bnb-4bit\",\n",
         "    \"unsloth/codellama-34b-bnb-4bit\",\n",
         "    \"unsloth/tinyllama-bnb-4bit\",\n",
-        "]\n",
+        "] # More models at https://huggingface.co/unsloth\n",
         "\n",
         "model, tokenizer = FastLanguageModel.from_pretrained(\n",
-        "    model_name = \"unsloth/mistral-7b-bnb-4bit\", # Choose ANY! eg mistralai/Mistral-7B-Instruct-v0.2\n",
+        "    model_name = \"unsloth/mistral-7b-bnb-4bit\", # Choose ANY! eg teknium/OpenHermes-2.5-Mistral-7B\n",
         "    max_seq_length = max_seq_length,\n",
         "    dtype = dtype,\n",
         "    load_in_4bit = load_in_4bit,\n",
@@ -348,7 +347,11 @@
         "\n",
         "**[NOTE]** To train only on completions (ignoring the user's input) read TRL's docs [here](https://huggingface.co/docs/trl/sft_trainer#train-on-completions-only).\n",
         "\n",
-        "**[NOTE]** Remember to add the **EOS_TOKEN** to the tokenized output!! Otherwise you'll get infinite generations!"
+        "**[NOTE]** Remember to add the **EOS_TOKEN** to the tokenized output!! Otherwise you'll get infinite generations!\n",
+        "\n",
+        "If you want to use the `ChatML` template for ShareGPT datasets, try our conversational [notebook](https://colab.research.google.com/drive/1Aau3lgPzeZKQ-98h69CCu1UJcvIBLmy2?usp=sharing).\n",
+        "\n",
+        "For text completions like novel writing, try this [notebook](https://colab.research.google.com/drive/1ef-tab5bhkvWmBOObepl1WgJvfvSzn5Q?usp=sharing)."
       ],
       "metadata": {
         "id": "vITh0KVJ10qX"
@@ -656,6 +659,7 @@
         "\n",
         "trainer = SFTTrainer(\n",
         "    model = model,\n",
+        "    tokenizer = tokenizer,\n",
         "    train_dataset = dataset,\n",
         "    dataset_text_field = \"text\",\n",
         "    max_seq_length = max_seq_length,\n",
@@ -1054,7 +1058,7 @@
       "cell_type": "code",
       "source": [
         "# alpaca_prompt = Copied from above\n",
-        "\n",
+        "FastLanguageModel.for_inference(model) # Enable native 2x faster inference\n",
         "inputs = tokenizer(\n",
         "[\n",
         "    alpaca_prompt.format(\n",
@@ -1062,7 +1066,7 @@
         "        \"1, 1, 2, 3, 5, 8\", # input\n",
         "        \"\", # output - leave this blank for generation!\n",
         "    )\n",
-        "]*1, return_tensors = \"pt\").to(\"cuda\")\n",
+        "], return_tensors = \"pt\").to(\"cuda\")\n",
         "\n",
         "outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True)\n",
         "tokenizer.batch_decode(outputs)"
@@ -1108,7 +1112,7 @@
       "cell_type": "code",
       "source": [
         "# alpaca_prompt = Copied from above\n",
-        "\n",
+        "FastLanguageModel.for_inference(model) # Enable native 2x faster inference\n",
         "inputs = tokenizer(\n",
         "[\n",
         "    alpaca_prompt.format(\n",
@@ -1116,7 +1120,7 @@
         "        \"1, 1, 2, 3, 5, 8\", # input\n",
         "        \"\", # output - leave this blank for generation!\n",
         "    )\n",
-        "]*1, return_tensors = \"pt\").to(\"cuda\")\n",
+        "], return_tensors = \"pt\").to(\"cuda\")\n",
         "\n",
         "from transformers import TextStreamer\n",
         "text_streamer = TextStreamer(tokenizer)\n",
@@ -1201,17 +1205,9 @@
         "        dtype = dtype,\n",
         "        load_in_4bit = load_in_4bit,\n",
         "    )\n",
+        "    FastLanguageModel.for_inference(model) # Enable native 2x faster inference\n",
         "\n",
-        "alpaca_prompt = \"\"\"Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n",
-        "\n",
-        "### Instruction:\n",
-        "{}\n",
-        "\n",
-        "### Input:\n",
-        "{}\n",
-        "\n",
-        "### Response:\n",
-        "{}\"\"\"\n",
+        "# alpaca_prompt = You MUST copy from above!\n",
         "\n",
         "inputs = tokenizer(\n",
         "[\n",
@@ -1220,7 +1216,7 @@
         "        \"\", # input\n",
         "        \"\", # output - leave this blank for generation!\n",
         "    )\n",
-        "]*1, return_tensors = \"pt\").to(\"cuda\")\n",
+        "], return_tensors = \"pt\").to(\"cuda\")\n",
         "\n",
         "outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True)\n",
         "tokenizer.batch_decode(outputs)"
@@ -1256,7 +1252,7 @@
     {
       "cell_type": "markdown",
       "source": [
-        "You can also use Hugging Face's `AutoModelForPeftCausalLM`"
+        "You can also use Hugging Face's `AutoModelForPeftCausalLM`. Only use this if you do not have `unsloth` installed. It can be hopelessly slow, since `4bit` model downloading is not supported, and Unsloth's **inference is 2x faster**."
       ],
       "metadata": {
         "id": "QQMjaNrjsU5_"
@@ -1266,6 +1262,7 @@
       "cell_type": "code",
       "source": [
         "if False:\n",
+        "    # I highly do NOT suggest - use Unsloth if possible\n",
         "    from peft import AutoModelForPeftCausalLM\n",
         "    from transformers import AutoTokenizer\n",
         "    model = AutoModelForPeftCausalLM.from_pretrained(\n",
@@ -1285,7 +1282,7 @@
       "source": [
         "### Saving to float16 for VLLM\n",
         "\n",
-        "We also support saving to `float16` directly. Select `merged_16bit` for float16 or `merged_4bit` for int4. We also allow `lora` adapters as a fallback. Use `push_to_hub_merged` to upload to your Hugging Face account!"
+        "We also support saving to `float16` directly. Select `merged_16bit` for float16 or `merged_4bit` for int4. We also allow `lora` adapters as a fallback. Use `push_to_hub_merged` to upload to your Hugging Face account! You can go to https://huggingface.co/settings/tokens for your personal tokens."
       ],
       "metadata": {
         "id": "f422JgM9sdVT"
@@ -1295,16 +1292,16 @@
       "cell_type": "code",
       "source": [
         "# Merge to 16bit\n",
-        "if False: model.save_pretrained_merged(\"x\", tokenizer, save_method = \"merged_16bit\",)\n",
-        "if False: model.push_to_hub_merged(\"hf_user/x\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
+        "if False: model.save_pretrained_merged(\"model\", tokenizer, save_method = \"merged_16bit\",)\n",
+        "if False: model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
         "\n",
         "# Merge to 4bit\n",
-        "if False: model.save_pretrained_merged(\"x\", tokenizer, save_method = \"merged_4bit\",)\n",
-        "if False: model.push_to_hub_merged(\"hf_user/x\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
+        "if False: model.save_pretrained_merged(\"model\", tokenizer, save_method = \"merged_4bit\",)\n",
+        "if False: model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
         "\n",
         "# Just LoRA adapters\n",
-        "if False: model.save_pretrained_merged(\"x\", tokenizer, save_method = \"lora\",)\n",
-        "if False: model.push_to_hub_merged(\"hf_user/x\", tokenizer, save_method = \"lora\", token = \"\")"
+        "if False: model.save_pretrained_merged(\"model\", tokenizer, save_method = \"lora\",)\n",
+        "if False: model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"lora\", token = \"\")"
       ],
       "metadata": {
         "id": "iHjt_SMYsd3P"
@@ -1316,7 +1313,12 @@
       "cell_type": "markdown",
       "source": [
         "### GGUF / llama.cpp Conversion\n",
-        "To save to `GGUF` / `llama.cpp`, we support it natively now! We clone `llama.cpp` and we default save it to `q8_0`. We allow all methods like `q4_k_m`. Use `save_pretrained_gguf` for local saving and `push_to_hub_gguf` for uploading to HF."
+        "To save to `GGUF` / `llama.cpp`, we support it natively now! We clone `llama.cpp` and we default save it to `q8_0`. We allow all methods like `q4_k_m`. Use `save_pretrained_gguf` for local saving and `push_to_hub_gguf` for uploading to HF.\n",
+        "\n",
+        "Some supported quant methods (full list on our [Wiki page](https://github.com/unslothai/unsloth/wiki#gguf-quantization-options)):\n",
+        "* `q8_0` - Fast conversion. High resource use, but generally acceptable.\n",
+        "* `q4_k_m` - Recommended. Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q4_K.\n",
+        "* `q5_k_m` - Recommended. Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q5_K."
       ],
       "metadata": {
         "id": "TCv4vXHd61i7"
@@ -1326,16 +1328,16 @@
       "cell_type": "code",
       "source": [
         "# Save to 8bit Q8_0\n",
-        "if False: model.save_pretrained_gguf(\"x\", tokenizer,)\n",
-        "if False: model.push_to_hub_gguf(\"hf_user/x\", tokenizer, token = \"\")\n",
+        "if False: model.save_pretrained_gguf(\"model\", tokenizer,)\n",
+        "if False: model.push_to_hub_gguf(\"hf/model\", tokenizer, token = \"\")\n",
         "\n",
         "# Save to 16bit GGUF\n",
-        "if False: model.save_pretrained_gguf(\"x\", tokenizer, quantization_method = \"f16\")\n",
-        "if False: model.push_to_hub_gguf(\"hf_user/x\", tokenizer, quantization_method = \"f16\", token = \"\")\n",
+        "if False: model.save_pretrained_gguf(\"model\", tokenizer, quantization_method = \"f16\")\n",
+        "if False: model.push_to_hub_gguf(\"hf/model\", tokenizer, quantization_method = \"f16\", token = \"\")\n",
         "\n",
         "# Save to q4_k_m GGUF\n",
-        "if False: model.save_pretrained_gguf(\"x\", tokenizer, quantization_method = \"q4_k_m\")\n",
-        "if False: model.push_to_hub_gguf(\"hf_user/x\", tokenizer, quantization_method = \"q4_k_m\", token = \"\")"
+        "if False: model.save_pretrained_gguf(\"model\", tokenizer, quantization_method = \"q4_k_m\")\n",
+        "if False: model.push_to_hub_gguf(\"hf/model\", tokenizer, quantization_method = \"q4_k_m\", token = \"\")"
       ],
       "metadata": {
         "id": "FqfebeAdT073"
@@ -1346,7 +1348,7 @@
     {
       "cell_type": "markdown",
       "source": [
-        "Now, use the `x.gguf` file or `x-unsloth-Q4_K_M.gguf` file in `llama.cpp` or a UI based system like `GPT4All`. You can install GPT4All by going [here](https://gpt4all.io/index.html)."
+        "Now, use the `model-unsloth.gguf` file or `model-unsloth-Q4_K_M.gguf` file in `llama.cpp` or a UI based system like `GPT4All`. You can install GPT4All by going [here](https://gpt4all.io/index.html)."
       ],
       "metadata": {
         "id": "bDp0zNpwe6U_"
@@ -1362,8 +1364,10 @@
         "2. Llama 7b 2x faster [free Colab](https://colab.research.google.com/drive/1lBzz5KeZJKXjvivbYvmGarix9Ao6Wxe5?usp=sharing)\n",
         "3. TinyLlama 4x faster full Alpaca 52K in 1 hour [free Colab](https://colab.research.google.com/drive/1AZghoNBQaMDgWJpi4RbffGM1h6raLUj9?usp=sharing)\n",
         "4. CodeLlama 34b 2x faster [A100 on Colab](https://colab.research.google.com/drive/1y7A0AxE3y8gdj4AVkl2aZX47Xu3P1wJT?usp=sharing)\n",
-        "5. Llama 7b [free Kaggle](https://www.kaggle.com/danielhanchen/unsloth-alpaca-t4-ddp)\n",
+        "5. Mistral 7b [free Kaggle version](https://www.kaggle.com/code/danielhanchen/kaggle-mistral-7b-unsloth-notebook)\n",
         "6. We also did a [blog](https://huggingface.co/blog/unsloth-trl) with 🤗 HuggingFace, and we're in the TRL [docs](https://huggingface.co/docs/trl/main/en/sft_trainer#accelerate-fine-tuning-2x-using-unsloth)!\n",
+        "7. `ChatML` for ShareGPT datasets, [conversational notebook](https://colab.research.google.com/drive/1Aau3lgPzeZKQ-98h69CCu1UJcvIBLmy2?usp=sharing)\n",
+        "8. Text completions like novel writing [notebook](https://colab.research.google.com/drive/1ef-tab5bhkvWmBOObepl1WgJvfvSzn5Q?usp=sharing)\n",
         "\n",
         "<div class=\"align-center\">\n",
         "  <a href=\"https://github.com/unslothai/unsloth\"><img src=\"https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png\" width=\"115\"></a>\n",