ep command-line interface can inspect evaluation runs locally, upload evaluators, and create reinforcement fine-tuning jobs on Fireworks.
Global Options
These options can be used with any command:boolean
default:"false"
Enable verbose logging (Aliases:
-v)Fireworks API server hostname or URL (e.g., dev.api.fireworks.ai or https://dev.api.fireworks.ai)
Commands
ep logs
Serve logs with file watching and real-time updates
number
default:"8000"
Port to bind to (default: 8000)
boolean
default:"false"
Enable debug mode
boolean
default:"false"
Disable Elasticsearch setup
boolean
default:"false"
Use env vars for Elasticsearch config (requires ELASTICSEARCH_URL, ELASTICSEARCH_API_KEY, ELASTICSEARCH_INDEX_NAME)
boolean
default:"false"
Force Fireworks tracing backend for logs UI (overrides env auto-detection)
boolean
default:"false"
Force Elasticsearch backend for logs UI (overrides env auto-detection)
ep upload
Scan for evaluation tests, select, and upload as Fireworks evaluators
string
default:"."
Path to search for evaluation tests (default: current directory)
Entrypoint of evaluation test to upload (module:function or path::function). For multiple, separate by commas.
boolean
default:"false"
Non-interactive: upload all discovered evaluation tests (Aliases:
-y)Path to .env file containing secrets to upload (default: .env in current directory)
boolean
default:"false"
Overwrite existing evaluator with the same ID
string
Default dataset to use with this evaluator (Aliases:
--default-dataset)string
Description for evaluator (Aliases:
--description)string
Display name for evaluator (defaults to ID) (Aliases:
--name, --display-name)string
Pytest-style entrypoint (e.g., test_file.py::test_func). Auto-detected if not provided. (Aliases:
--entry-point)string
Requirements for evaluator (auto-detected from requirements.txt if not provided) (Aliases:
--requirements)string
Evaluator ID to use (if multiple selections, a numeric suffix is appended) (Aliases:
--id)ep create rft
Create a Reinforcement Fine-tuning Job on Fireworks
boolean
default:"false"
Non-interactive mode (Aliases:
-y)boolean
default:"false"
Print planned SDK call without sending
boolean
default:"false"
Overwrite existing evaluator with the same ID
boolean
default:"false"
Skip local dataset/evaluator validation
boolean
default:"false"
Ignore Dockerfile even if present; run pytest on host during evaluator validation
string
default:""
Extra flags to pass to ‘docker build’ when validating evaluator (quoted string, e.g. “—no-cache —pull —progress=plain”)
string
default:""
Extra flags to pass to ‘docker run’ when validating evaluator (quoted string, e.g. “—env-file .env —memory=8g”)
Path to .env file containing secrets to upload to Fireworks (default: .env in project root)
The source reinforcement fine-tuning job to copy configuration from. If other flags are set, they will override the source job’s configuration.
boolean
default:"false"
If set, only errors will be printed.
string
The name of the dataset used for training.
string
The evaluator resource name to use for RLOR fine-tuning job.
string
ID of the reinforcement fine-tuning job, a random UUID will be generated if not specified. (Aliases:
--job-id)number
Data chunking for rollout, default size 200, enabled when dataset > 300. Valid range is 1-10,000.
boolean
default:"false"
Whether to auto-carve the dataset for eval.
string
The name of a separate dataset to use for evaluation.
string
Additional parameters for the inference request as a JSON string. For example:
”{“stop”: [“\n”]}”. (Aliases:
--extra-body)number
Maximum number of tokens to generate per response. (Aliases:
--max-output-tokens)number
Number of response candidates to generate per input. (Aliases:
--response-candidates-count)number
Sampling temperature, typically between 0 and 2. (Aliases:
--temperature)number
Top-k sampling parameter, limits the token selection to the top k tokens. (Aliases:
--top-k)number
Top-p sampling parameter, typically between 0 and 1. (Aliases:
--top-p)number
KL coefficient (beta) override for GRPO-like methods. If unset, the trainer
default is used. (Aliases:
--rl-kl-beta, --kl-beta)Literal
RL loss method for underlying trainers. One of {grpo,dapo}. (Aliases:
--rl-loss-method, --method)string
The MCP server resource name to use for the reinforcement fine-tuning job. (Optional)
number
The number of nodes to use for the fine-tuning job. If not specified, the default is 1. (Aliases:
--nodes)string
The name of the base model to be fine-tuned Only one of ‘base_model’ or
‘warm_start_from’ should be specified. (Aliases:
--base-model)number
The maximum packed number of tokens per batch for training in sequence packing. (Aliases:
--batch-size)number
The number of epochs to train for. (Aliases:
--epochs)number
The number of batches to accumulate gradients before updating the model parameters. The effective batch size will be batch-size multiplied by this value. (Aliases:
--gradient-accumulation-steps)number
The learning rate used for training. (Aliases:
--learning-rate)number
The number of learning rate warmup steps for the reinforcement fine-tuning job. (Aliases:
--learning-rate-warmup-steps)number
The rank of the LoRA layers. (Aliases:
--lora-rank)number
The maximum context length to use with the model. (Aliases:
--max-context-length)string
The model ID to be assigned to the resulting fine-tuned model.If not specified, the job ID will be used. (Aliases:
--output-model)string
The PEFT addon model in Fireworks format to be fine-tuned from Only one of
‘base_model’ or ‘warm_start_from’ should be specified. (Aliases:
--warm-start-from)string
The API key for the wandb service. (Aliases:
--wandb-api-key, --api-key)boolean
default:"false"
Whether to enable wandb logging. (Aliases:
--wandb, --enabled)string
The entity name for the wandb service. (Aliases:
--wandb-entity, --entity)string
The project name for the wandb service. (Aliases:
--wandb-project, --project)ep local-test
Select an evaluation test and run it locally. If a Dockerfile exists, build and run via Docker; otherwise run on host.
Entrypoint to run (path::function or path). If not provided, a selector will be shown (unless —yes).
boolean
default:"false"
Ignore Dockerfile even if present; run pytest on host
boolean
default:"false"
Non-interactive: if multiple tests exist and no —entry, fails with guidance (Aliases:
-y)string
default:""
Extra flags to pass to ‘docker build’ (quoted string, e.g. “—no-cache —pull —progress=plain”)
string
default:""
Extra flags to pass to ‘docker run’ (quoted string, e.g. “—env-file .env —memory=8g”)