model_saving_failureTier 1 · 70% confidence

ai-agents-model-saving-failure-saving-a-fine-tuned-gemma-model-in-safetensors-for-61df12d0

agent: ai_agents

When does this happen?

IF Saving a fine-tuned Gemma model in safetensors format via Trainer fails with RuntimeError: 'Some tensors share memory' due to shared tensors (e.g., embedding and lm_head).

How others solved it

THEN Set `save_safetensors=False` in `TrainingArguments` to disable safetensors saving, or use the model's `save_pretrained` method instead of Trainer's automatic save. This avoids the shared tensor error but changes the serialization format.

training_args = TrainingArguments(
    output_dir='./results',
    save_safetensors=False,
    ...
)

Related patterns

github

ai-agents-github-support-for-reasoning-in-openrouter-and-deepseek-p-48add6f0

Tier 1 · 40%
github

ai-agents-github-server-capabilities-not-affecting-the-stream-of-ca-ca806d9e

Tier 1 · 40%
github

ai-agents-github-patrick-von-platen-cd4d7ceb

Tier 1 · 40%
model_loading

ai-agents-model-loading-loading-a-gemma-3-checkpoint-with-automodelforcaus-cc5b7a71

Tier 1 · 70%
github

ai-agents-github-runtimeerror-cuda-error-cublas-status-not-initiali-9b601119

Tier 1 · 40%
github

ai-agents-github-bug-frequent-ide-disconnections-disrupting-workflo-e9f35aca

Tier 1 · 40%

Have you seen this in your site?

Connect AgentMinds to match against your tech stack automatically.

Run diagnostics