From 4ba22789dafe476b299c6b1c9d58f65cecc3021a Mon Sep 17 00:00:00 2001 From: 1stprinciple Date: Fri, 30 Jan 2026 01:23:24 +0100 Subject: [PATCH] add gspo-token to loss_config.method --- eval_protocol/cli.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/eval_protocol/cli.py b/eval_protocol/cli.py index 46e3c376..ade6df0e 100644 --- a/eval_protocol/cli.py +++ b/eval_protocol/cli.py @@ -200,7 +200,7 @@ def _configure_parser(parser: argparse.ArgumentParser) -> argparse.ArgumentParse "training_config.gradient_accumulation_steps": "The number of batches to accumulate gradients before updating the model parameters. The effective batch size will be batch-size multiplied by this value.", "training_config.learning_rate_warmup_steps": "The number of learning rate warmup steps for the reinforcement fine-tuning job.", "mcp_server": "The MCP server resource name to use for the reinforcement fine-tuning job. (Optional)", - "loss_config.method": "RL loss method for underlying trainers. One of {grpo,dapo}.", + "loss_config.method": "RL loss method for underlying trainers. One of {grpo,dapo,gspo-token}.", } create_rft_job_fn = Fireworks().reinforcement_fine_tuning_jobs.create