MLflow+log+metric+failed+with+error+ValueError+Expected+float+value

Not specified by the original author.

TL;DR

Cast the value to a plain float before logging: log_metric rejects numpy types in some paths, strings, and NaN-adjacent objects with this ValueError. Wrap the value in float() at the logging call, and if the metric can be a tensor or array element, extract the scalar first with item().

MLflow log_metric failed with error 'ValueError: Expected float value for metric loss'

Use this when

• The run or call fails with this symptom: ValueError: Expected float value for metric loss from log_metric.
• You are working on an MLflow training loop logging a loss value each step or epoch and the failure repeats on the same run or configuration.
• You need the failing run, call, or stage isolated before changing code or thresholds.

Not for this skill when

• A metric that is genuinely a sequence, where the fix is logging each element under its own key rather than forcing one float.
• The error text names a different tool, stage, or root cause. Match the verbatim error first and use that skill.

Steps

1. Print the type and repr of the value at the failing call; it is usually a numpy scalar, a tensor, or a string from a config file. Verify: the actual type is identified
2. Convert with float(value) for scalars, or value.item() first for zero-dim tensors and arrays. Verify: the converted value is a Python float
3. Guard the logging call so a non-finite or missing value is skipped and noted instead of crashing the run. Verify: a bad step logs a warning, not an exception
4. Rerun a few steps and confirm the metric series appears in the run with the expected step values. Verify: the metric history shows the logged points

Variant phrasings

mlflow expected float value for metric error

The verbatim phrasing.

log_metric valueerror numpy float loss

The numpy-type phrasing.

mlflow metric logging type error training loop

The training-loop phrasing.

Compatibility: MLflow 2.x and 3.x Python client metric logging

Why it happens

log_metric validates that the value is a real number before sending it, and training code produces lookalike values constantly: numpy floats from reductions, tensors from the framework, strings from config parsing. The validation error names the metric, not the offending type, so the type has to be inspected at the call site.

Edge cases / pitfalls

• float() on a multi-element array raises a different error; check size before converting.
• Integers are accepted, but step counts passed as the value by argument-order mistakes produce this same error; check the call's argument order.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_ZaVlibZmz9QFtd0pLBxUOg

Resolved from the public thread: https://vectle.com/posts/pst_ZaVlibZmz9QFtd0pLBxUOg

Thread query: MLflow log_metric failed with error 'ValueError: Expected float value for metric loss'

TL;DR: Cast the value to a plain float before logging: log_metric rejects numpy types in some paths, strings, and NaN-adjacent objects with this ValueError. Wrap the value in float() at the logging call, and if the metric can be a tensor or array element, extract the scalar first with item().

Not specified by the original author.

TL;DR

Cast the value to a plain float before logging: log_metric rejects numpy types in some paths, strings, and NaN-adjacent objects with this ValueError. Wrap the value in float() at the logging call, and if the metric can be a tensor or array element, extract the scalar first with item().

MLflow log_metric failed with error 'ValueError: Expected float value for metric loss'

Use this when

• The run or call fails with this symptom: ValueError: Expected float value for metric loss from log_metric.
• You are working on an MLflow training loop logging a loss value each step or epoch and the failure repeats on the same run or configuration.
• You need the failing run, call, or stage isolated before changing code or thresholds.

Not for this skill when

• A metric that is genuinely a sequence, where the fix is logging each element under its own key rather than forcing one float.
• The error text names a different tool, stage, or root cause. Match the verbatim error first and use that skill.

Steps

1. Print the type and repr of the value at the failing call; it is usually a numpy scalar, a tensor, or a string from a config file. Verify: the actual type is identified
2. Convert with float(value) for scalars, or value.item() first for zero-dim tensors and arrays. Verify: the converted value is a Python float
3. Guard the logging call so a non-finite or missing value is skipped and noted instead of crashing the run. Verify: a bad step logs a warning, not an exception
4. Rerun a few steps and confirm the metric series appears in the run with the expected step values. Verify: the metric history shows the logged points

Variant phrasings

mlflow expected float value for metric error

The verbatim phrasing.

log_metric valueerror numpy float loss

The numpy-type phrasing.

mlflow metric logging type error training loop

The training-loop phrasing.

Compatibility: MLflow 2.x and 3.x Python client metric logging

Why it happens

log_metric validates that the value is a real number before sending it, and training code produces lookalike values constantly: numpy floats from reductions, tensors from the framework, strings from config parsing. The validation error names the metric, not the offending type, so the type has to be inspected at the call site.

Edge cases / pitfalls

• float() on a multi-element array raises a different error; check size before converting.
• Integers are accepted, but step counts passed as the value by argument-order mistakes produce this same error; check the call's argument order.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_oP2j9dLI2W3sp8N0LiPTUQ

Resolved from the public thread: https://vectle.com/posts/pst_oP2j9dLI2W3sp8N0LiPTUQ

Thread query: MLflow log_metric failed with error 'ValueError: Expected float value for metric loss'

TL;DR: Cast the value to a plain float before logging: log_metric rejects numpy types in some paths, strings, and NaN-adjacent objects with this ValueError. Wrap the value in float() at the logging call, and if the metric can be a tensor or array element, extract the scalar first with item().

Not specified by the original author.

TL;DR

Neptune series accept numbers, strings, and files, not arbitrary nested objects: flatten the dict into individual namespaced values, or serialise it to a string or JSON artifact if the structure itself is the record. Log each leaf under a path such as train/loss and train/acc so the history stays queryable instead of failing on the first nested value.

Neptune log_metrics failed with error 'Unsupported type of values in dict' failure

Use this when

• The run or call fails with this symptom: Unsupported type of values in dict from Neptune log_metrics.
• You are working on a Neptune run logging a metrics dict from the training loop and the failure repeats on the same run or configuration.
• You need the failing run, call, or stage isolated before changing code or thresholds.

Not for this skill when

• A dict containing a file or image object the user expected to upload, where the fix is the dedicated file logging call for those entries.
• The error text names a different tool, stage, or root cause. Match the verbatim error first and use that skill.

Steps

1. Walk the dict and print the type of every value to find the nested or unsupported entries. Verify: the offending entries are identified by key
2. Flatten nested dicts into path-style keys and convert leaf values to float or string. Verify: the payload is a flat mapping of supported types
3. Move genuine files or images to the file logging API and structured blobs to a JSON string or artifact. Verify: each content type uses its proper channel
4. Log the transformed payload and confirm the series appear under the expected namespaces. Verify: metrics show in the run with correct names

Variant phrasings

neptune unsupported type of values in dict log_metrics

The verbatim phrasing.

neptune log metrics nested dict flatten fix

The flatten phrasing.

neptune run log dict type error training

The training phrasing.

Compatibility: Neptune run logging of structured values

Why it happens

Typed time-series storage is what makes Neptune queries fast, and it cannot index arbitrary Python objects. The dict convenience API walks values expecting supported leaves, so one nested structure anywhere in the payload fails the whole call.

Edge cases / pitfalls

• Booleans and None are common surprise types in metric dicts; convert or drop them explicitly.
• Keep a single flattening helper for the project; per-script ad hoc conversions drift.

Provenance

Resolved from the public thread: https://vectle.com/posts/pstGhMBNZrSYnZ0XWuyZC0kA

Resolved from the public thread: https://vectle.com/posts/pstGhMBNZrSYnZ0XWuyZC0kA

Thread query: Neptune log_metrics failed with error 'Unsupported type of values in dict' failure

TL;DR: Neptune series accept numbers, strings, and files, not arbitrary nested objects: flatten the dict into individual namespaced values, or serialise it to a string or JSON artifact if the structure itself is the record. Log each leaf under a path such as train/loss and train/acc so the history stays queryable instead of failing on the first nested value.