1.7b-MixtureVitae-300BT-v1-16k-WithChatTemplate

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

opensci_full_sft_fsdp_offload

This model is a fine-tuned version of ontocord/1.7b-MixtureVitae-300BT-v1 on the smoltalk, the longalign and the sealong datasets. The model extends the context length of base model from 4096 to 16384 tokens.

NOTE: Because the post-training data is no longer strictly public domain, cc-by and .gov, please do not depend on these constraints on the model output. This is an experimental model.

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 64
• gradient_accumulation_steps: 2
• total_train_batch_size: 512
• total_eval_batch_size: 512
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 1.0

Training results

Framework versions

• Transformers 4.55.0
• Pytorch 2.6.0+cu124
• Datasets 4.0.0
• Tokenizers 0.21.4