Skip to content

Change the vllm saving cells #147

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Etherll wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Change the vllm saving cells #147

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
16 changes: 8 additions & 8 deletions nb/Advanced_Llama3_1_(3B)_GRPO_LoRA.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9703,20 +9703,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
Comment on lines 9705 to +9719

@gemini-code-assist gemini-code-assist Bot Dec 19, 2025

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

While the change to use unique names is good, this cell has a lot of duplicated 'magic strings' for model paths. This can be improved for better maintainability by defining constants at the top of the cell and reusing them. This would also make it easier to update paths in the future. This principle applies to all similar saving cells across the modified notebooks.

For example, you could refactor this cell's source code to look something like this:

# Define constants for model paths
MODEL_MERGED_16BIT_PATH = "model-merged"
HF_HUB_MERGED_16BIT_PATH = f"hf/{MODEL_MERGED_16BIT_PATH}"

MODEL_MERGED_4BIT_PATH = "model-merged-4bit"
HF_HUB_MERGED_4BIT_PATH = f"hf/{MODEL_MERGED_4BIT_PATH}"

LORA_MODEL_PATH = "lora_model"
HF_HUB_LORA_PATH = f"hf/{LORA_MODEL_PATH}"

# Merge to 16bit
if False: model.save_pretrained_merged(MODEL_MERGED_16BIT_PATH, tokenizer, save_method = "merged_16bit",)
if False: model.push_to_hub_merged(HF_HUB_MERGED_16BIT_PATH, tokenizer, save_method = "merged_16bit", token = "")

# Merge to 4bit
if False: model.save_pretrained_merged(MODEL_MERGED_4BIT_PATH, tokenizer, save_method = "merged_4bit",)
if False: model.push_to_hub_merged(HF_HUB_MERGED_4BIT_PATH, tokenizer, save_method = "merged_4bit", token = "")

# Just LoRA adapters
if False:
    model.save_pretrained(LORA_MODEL_PATH)
    tokenizer.save_pretrained(LORA_MODEL_PATH)
if False:
    model.push_to_hub(HF_HUB_LORA_PATH, token = "")
    tokenizer.push_to_hub(HF_HUB_LORA_PATH, token = "")
References
  1. Magic strings (and numbers) should be replaced by named constants to improve readability and maintainability. This makes the code easier to understand and modify, as the value is defined in one place.

]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/Advanced_Llama3_2_(3B)_GRPO_LoRA.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12647,20 +12647,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/CodeForces-cot-Finetune_for_Reasoning_on_CodeForces.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1473,20 +1473,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/CodeGemma_(7B)-Conversational.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -824,20 +824,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/DeepSeek_R1_0528_Qwen3_(8B)_GRPO.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5022,20 +5022,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
4 changes: 2 additions & 2 deletions nb/Deepseek_OCR_(3B)-Eval.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1647,10 +1647,10 @@
"# Select ONLY 1 to save! (Both not needed!)\n",
"\n",
"# Save locally to 16bit\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune\", tokenizer,)\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune-merged\", tokenizer,)\n",
"\n",
"# To export and save to your Hugging Face account\n",
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune\", tokenizer, token = \"PUT_HERE\")"
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune-merged\", tokenizer, token = \"PUT_HERE\")"
]
},
{
Expand Down
4 changes: 2 additions & 2 deletions nb/Deepseek_OCR_(3B)-Evaluation.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1961,10 +1961,10 @@
"# Select ONLY 1 to save! (Both not needed!)\n",
"\n",
"# Save locally to 16bit\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune\", tokenizer,)\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune-merged\", tokenizer,)\n",
"\n",
"# To export and save to your Hugging Face account\n",
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune\", tokenizer, token = \"PUT_HERE\")"
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune-merged\", tokenizer, token = \"PUT_HERE\")"
]
},
{
Expand Down
4 changes: 2 additions & 2 deletions nb/Deepseek_OCR_(3B).ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1647,10 +1647,10 @@
"# Select ONLY 1 to save! (Both not needed!)\n",
"\n",
"# Save locally to 16bit\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune\", tokenizer,)\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune-merged\", tokenizer,)\n",
"\n",
"# To export and save to your Hugging Face account\n",
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune\", tokenizer, token = \"PUT_HERE\")"
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune-merged\", tokenizer, token = \"PUT_HERE\")"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/ERNIE_4_5_21B_A3B_PT-Conversational.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1692,23 +1692,23 @@
"source": [
"# Merge to 16bit\n",
"if False:\n",
" model.save_pretrained_merged(\"model\", tokenizer, save_method = \"merged_16bit\",)\n",
" model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
" model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\n",
"if False:\n",
" model.save_pretrained_merged(\"model\", tokenizer, save_method = \"merged_4bit\",)\n",
" model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub_merged(\"hf/model\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
" model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
4 changes: 2 additions & 2 deletions nb/ERNIE_4_5_VL_28B_A3B_PT_Vision.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2089,10 +2089,10 @@
"# Select ONLY 1 to save! (Both not needed!)\n",
"\n",
"# Save locally to 16bit\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune\", tokenizer,)\n",
"if False: model.save_pretrained_merged(\"unsloth_finetune-merged\", tokenizer,)\n",
"\n",
"# To export and save to your Hugging Face account\n",
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune\", tokenizer, token = \"PUT_HERE\")"
"if False: model.push_to_hub_merged(\"YOUR_USERNAME/unsloth_finetune-merged\", tokenizer, token = \"PUT_HERE\")"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/Falcon_H1_(0.5B)-Alpaca.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -989,20 +989,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/FunctionGemma_(270M)-Mobile-Actions.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1689,23 +1689,23 @@
"source": [
"# Merge to 16bit\n",
"if False:\n",
" model.save_pretrained_merged(\"functiongemma-finetune\", tokenizer, save_method = \"merged_16bit\")\n",
" model.save_pretrained_merged(\"functiongemma-finetune-merged\", tokenizer, save_method = \"merged_16bit\")\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\n",
"if False:\n",
" model.save_pretrained_merged(\"functiongemma-finetune\", tokenizer, save_method = \"merged_4bit\",)\n",
" model.save_pretrained_merged(\"functiongemma-finetune-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"functiongemma-finetune\")\n",
" tokenizer.save_pretrained(\"functiongemma-finetune\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub(\"hf/functiongemma-finetune\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/functiongemma-finetune\", token = \"\")"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")"
],
"metadata": {
"id": "Ql1Nqa76y9af"
Expand Down
16 changes: 8 additions & 8 deletions nb/FunctionGemma_(270M).ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2227,23 +2227,23 @@
"source": [
"# Merge to 16bit\n",
"if False:\n",
" model.save_pretrained_merged(\"functiongemma-finetune\", tokenizer, save_method = \"merged_16bit\")\n",
" model.save_pretrained_merged(\"functiongemma-finetune-merged\", tokenizer, save_method = \"merged_16bit\")\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\n",
"if False:\n",
" model.save_pretrained_merged(\"functiongemma-finetune\", tokenizer, save_method = \"merged_4bit\",)\n",
" model.save_pretrained_merged(\"functiongemma-finetune-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
" model.push_to_hub_merged(\"hf/functiongemma-finetune-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"functiongemma-finetune\")\n",
" tokenizer.save_pretrained(\"functiongemma-finetune\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False: # Pushing to HF Hub\n",
" model.push_to_hub(\"hf/functiongemma-finetune\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/functiongemma-finetune\", token = \"\")"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")"
]
},
{
Expand Down
16 changes: 8 additions & 8 deletions nb/Gemma2_(2B)-Alpaca.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1197,20 +1197,20 @@
"outputs": [],
"source": [
"# Merge to 16bit\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",
"if False: model.save_pretrained_merged(\"model-merged\", tokenizer, save_method = \"merged_16bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged\", tokenizer, save_method = \"merged_16bit\", token = \"\")\n",
"\n",
"# Merge to 4bit\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",
"if False: model.save_pretrained_merged(\"model-merged-4bit\", tokenizer, save_method = \"merged_4bit\",)\n",
"if False: model.push_to_hub_merged(\"hf/model-merged-4bit\", tokenizer, save_method = \"merged_4bit\", token = \"\")\n",
"\n",
"# Just LoRA adapters\n",
"if False:\n",
" model.save_pretrained(\"model\")\n",
" tokenizer.save_pretrained(\"model\")\n",
" model.save_pretrained(\"lora_model\")\n",
" tokenizer.save_pretrained(\"lora_model\")\n",
"if False:\n",
" model.push_to_hub(\"hf/model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/model\", token = \"\")\n"
" model.push_to_hub(\"hf/lora_model\", token = \"\")\n",
" tokenizer.push_to_hub(\"hf/lora_model\", token = \"\")\n"
]
},
{
Expand Down
Loading