Spaces:
Running
Running
adds optimizations for faster training
Browse files
config/train_gpt_oss_openhermes_fr_memory_optimized.py
CHANGED
@@ -56,7 +56,7 @@ config = GPTOSSEnhancedCustomConfig(
|
|
56 |
# MODEL CONFIGURATION - Memory Optimized for GPT-OSS
|
57 |
# ============================================================================
|
58 |
model_name="openai/gpt-oss-20b",
|
59 |
-
max_seq_length=
|
60 |
use_flash_attention=True, # Critical for memory efficiency
|
61 |
use_gradient_checkpointing=True, # Essential for memory optimization
|
62 |
|
@@ -115,9 +115,10 @@ config = GPTOSSEnhancedCustomConfig(
|
|
115 |
},
|
116 |
|
117 |
# Data loading optimized for throughput
|
118 |
-
dataloader_num_workers=
|
119 |
dataloader_pin_memory=True, # Pin memory for faster host->GPU copies
|
120 |
-
dataloader_prefetch_factor=
|
|
|
121 |
|
122 |
# Memory management optimizations
|
123 |
max_memory_per_gpu=None, # No explicit memory limit; use as much VRAM as available
|
@@ -197,6 +198,9 @@ config = GPTOSSEnhancedCustomConfig(
|
|
197 |
"min_lr": 2e-6, # Explicit absolute floor (matches min_lr above)
|
198 |
"warmup_steps": None, # Use warmup_ratio instead
|
199 |
},
|
|
|
|
|
|
|
200 |
|
201 |
# ============================================================================
|
202 |
# MONITORING & HUB INTEGRATION
|
|
|
56 |
# MODEL CONFIGURATION - Memory Optimized for GPT-OSS
|
57 |
# ============================================================================
|
58 |
model_name="openai/gpt-oss-20b",
|
59 |
+
max_seq_length=2048, # Shorter context speeds steps without reducing sample count
|
60 |
use_flash_attention=True, # Critical for memory efficiency
|
61 |
use_gradient_checkpointing=True, # Essential for memory optimization
|
62 |
|
|
|
115 |
},
|
116 |
|
117 |
# Data loading optimized for throughput
|
118 |
+
dataloader_num_workers=8, # More workers for faster loading
|
119 |
dataloader_pin_memory=True, # Pin memory for faster host->GPU copies
|
120 |
+
dataloader_prefetch_factor=2, # Slightly higher prefetch for throughput
|
121 |
+
dataset_num_proc=8, # Parallelize HF datasets map/filter
|
122 |
|
123 |
# Memory management optimizations
|
124 |
max_memory_per_gpu=None, # No explicit memory limit; use as much VRAM as available
|
|
|
198 |
"min_lr": 2e-6, # Explicit absolute floor (matches min_lr above)
|
199 |
"warmup_steps": None, # Use warmup_ratio instead
|
200 |
},
|
201 |
+
|
202 |
+
# Packing to increase token utilization per step (supported by TRL)
|
203 |
+
packing=True,
|
204 |
|
205 |
# ============================================================================
|
206 |
# MONITORING & HUB INTEGRATION
|
scripts/training/train_gpt_oss.py
CHANGED
@@ -210,6 +210,13 @@ def build_scheduler_kwargs(config):
|
|
210 |
def apply_dataset_filtering(dataset, config):
|
211 |
"""Apply filtering based on configuration"""
|
212 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
213 |
# Filter bad entries if specified
|
214 |
if getattr(config, 'filter_bad_entries', False):
|
215 |
bad_entry_field = getattr(config, 'bad_entry_field', 'bad_entry')
|
@@ -220,17 +227,23 @@ def apply_dataset_filtering(dataset, config):
|
|
220 |
|
221 |
# Filter out bad entries
|
222 |
if bad_entry_field in dataset.column_names:
|
223 |
-
|
|
|
|
|
224 |
print(f"Filtered {original_size - len(dataset)} bad entries")
|
225 |
|
226 |
# Filter out bad prompts
|
227 |
if bad_prompt_field in dataset.column_names:
|
228 |
-
|
|
|
|
|
229 |
print(f"Filtered bad prompts, remaining: {len(dataset)} examples")
|
230 |
|
231 |
# Filter out bad responses
|
232 |
if bad_response_field in dataset.column_names:
|
233 |
-
|
|
|
|
|
234 |
print(f"Filtered bad responses, remaining: {len(dataset)} examples")
|
235 |
|
236 |
# Apply length filtering
|
@@ -253,7 +266,7 @@ def apply_dataset_filtering(dataset, config):
|
|
253 |
return True
|
254 |
|
255 |
original_size = len(dataset)
|
256 |
-
dataset = dataset.filter(length_filter)
|
257 |
print(f"Length filtering: {original_size} -> {len(dataset)} examples")
|
258 |
|
259 |
# Apply sampling if specified
|
@@ -293,6 +306,13 @@ def format_gpt_oss_harmony_prompt(prompt: str) -> str:
|
|
293 |
def process_dataset_format(dataset, config):
|
294 |
"""Process dataset based on format configuration with exact GPT-OSS Harmony compliance"""
|
295 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
296 |
dataset_format = getattr(config, 'dataset_format', 'openhermes_fr')
|
297 |
input_field = getattr(config, 'input_field', 'prompt')
|
298 |
target_field = getattr(config, 'target_field', 'accepted_completion')
|
@@ -325,7 +345,7 @@ def process_dataset_format(dataset, config):
|
|
325 |
return {"prompt": prompt_val, "chosen": chosen_val, "rejected": rejected_val}
|
326 |
|
327 |
keep_cols = [c for c in ['prompt', 'chosen', 'rejected'] if c in dataset.column_names]
|
328 |
-
dataset = dataset.map(id_map, remove_columns=dataset.column_names if keep_cols else dataset.column_names)
|
329 |
return dataset
|
330 |
|
331 |
# Custom preference mapping via configured field names
|
@@ -341,7 +361,7 @@ def process_dataset_format(dataset, config):
|
|
341 |
return {"prompt": prompt_text, "chosen": chosen_text, "rejected": rejected_text}
|
342 |
return {"prompt": prompt_val, "chosen": chosen_val, "rejected": rejected_val}
|
343 |
|
344 |
-
dataset = dataset.map(to_pref, remove_columns=dataset.column_names)
|
345 |
return dataset
|
346 |
|
347 |
# If we reach here, we don't have required fields for DPO
|
@@ -371,7 +391,7 @@ def process_dataset_format(dataset, config):
|
|
371 |
"output": completion
|
372 |
}
|
373 |
|
374 |
-
dataset = dataset.map(format_openhermes_fr, remove_columns=dataset.column_names)
|
375 |
|
376 |
elif dataset_format == "messages":
|
377 |
# Process messages format (like HuggingFaceH4/Multilingual-Thinking)
|
@@ -416,7 +436,7 @@ def process_dataset_format(dataset, config):
|
|
416 |
|
417 |
return {"text": text}
|
418 |
|
419 |
-
dataset = dataset.map(format_messages, remove_columns=dataset.column_names)
|
420 |
|
421 |
elif dataset_format == "text":
|
422 |
# Process plain text format
|
@@ -427,7 +447,7 @@ def process_dataset_format(dataset, config):
|
|
427 |
text += "</s>"
|
428 |
return {"text": text}
|
429 |
|
430 |
-
dataset = dataset.map(format_text, remove_columns=dataset.column_names)
|
431 |
|
432 |
elif dataset_format == "custom":
|
433 |
# Custom format - user handles this in their config
|
@@ -652,6 +672,8 @@ def create_sft_config(config, output_dir):
|
|
652 |
"bf16": bf16,
|
653 |
# Some versions support tf32
|
654 |
"tf32": tf32 if 'tf32' in TrainingArguments.__init__.__code__.co_varnames else None,
|
|
|
|
|
655 |
# Regularization
|
656 |
"weight_decay": weight_decay,
|
657 |
"max_grad_norm": max_grad_norm,
|
@@ -828,6 +850,10 @@ def train_gpt_oss(config_path, experiment_name, output_dir, trackio_url, trainer
|
|
828 |
if "max_seq_length" in sft_params:
|
829 |
sft_kwargs["max_seq_length"] = getattr(config, 'max_seq_length', 2048)
|
830 |
|
|
|
|
|
|
|
|
|
831 |
# Remove any None values
|
832 |
sft_kwargs = {k: v for k, v in sft_kwargs.items() if v is not None}
|
833 |
|
|
|
210 |
def apply_dataset_filtering(dataset, config):
|
211 |
"""Apply filtering based on configuration"""
|
212 |
|
213 |
+
# Parallel workers for datasets ops
|
214 |
+
try:
|
215 |
+
import os as _os
|
216 |
+
num_proc = getattr(config, 'dataset_num_proc', None) or (_os.cpu_count() or 1)
|
217 |
+
except Exception:
|
218 |
+
num_proc = 1
|
219 |
+
|
220 |
# Filter bad entries if specified
|
221 |
if getattr(config, 'filter_bad_entries', False):
|
222 |
bad_entry_field = getattr(config, 'bad_entry_field', 'bad_entry')
|
|
|
227 |
|
228 |
# Filter out bad entries
|
229 |
if bad_entry_field in dataset.column_names:
|
230 |
+
def _keep_not_bad_entry(example, _field=bad_entry_field):
|
231 |
+
return not example.get(_field, False)
|
232 |
+
dataset = dataset.filter(_keep_not_bad_entry, num_proc=num_proc)
|
233 |
print(f"Filtered {original_size - len(dataset)} bad entries")
|
234 |
|
235 |
# Filter out bad prompts
|
236 |
if bad_prompt_field in dataset.column_names:
|
237 |
+
def _keep_not_bad_prompt(example, _field=bad_prompt_field):
|
238 |
+
return not example.get(_field, False)
|
239 |
+
dataset = dataset.filter(_keep_not_bad_prompt, num_proc=num_proc)
|
240 |
print(f"Filtered bad prompts, remaining: {len(dataset)} examples")
|
241 |
|
242 |
# Filter out bad responses
|
243 |
if bad_response_field in dataset.column_names:
|
244 |
+
def _keep_not_bad_response(example, _field=bad_response_field):
|
245 |
+
return not example.get(_field, False)
|
246 |
+
dataset = dataset.filter(_keep_not_bad_response, num_proc=num_proc)
|
247 |
print(f"Filtered bad responses, remaining: {len(dataset)} examples")
|
248 |
|
249 |
# Apply length filtering
|
|
|
266 |
return True
|
267 |
|
268 |
original_size = len(dataset)
|
269 |
+
dataset = dataset.filter(length_filter, num_proc=num_proc)
|
270 |
print(f"Length filtering: {original_size} -> {len(dataset)} examples")
|
271 |
|
272 |
# Apply sampling if specified
|
|
|
306 |
def process_dataset_format(dataset, config):
|
307 |
"""Process dataset based on format configuration with exact GPT-OSS Harmony compliance"""
|
308 |
|
309 |
+
# Parallel workers for datasets ops
|
310 |
+
try:
|
311 |
+
import os as _os
|
312 |
+
num_proc = getattr(config, 'dataset_num_proc', None) or (_os.cpu_count() or 1)
|
313 |
+
except Exception:
|
314 |
+
num_proc = 1
|
315 |
+
|
316 |
dataset_format = getattr(config, 'dataset_format', 'openhermes_fr')
|
317 |
input_field = getattr(config, 'input_field', 'prompt')
|
318 |
target_field = getattr(config, 'target_field', 'accepted_completion')
|
|
|
345 |
return {"prompt": prompt_val, "chosen": chosen_val, "rejected": rejected_val}
|
346 |
|
347 |
keep_cols = [c for c in ['prompt', 'chosen', 'rejected'] if c in dataset.column_names]
|
348 |
+
dataset = dataset.map(id_map, remove_columns=dataset.column_names if keep_cols else dataset.column_names, num_proc=num_proc)
|
349 |
return dataset
|
350 |
|
351 |
# Custom preference mapping via configured field names
|
|
|
361 |
return {"prompt": prompt_text, "chosen": chosen_text, "rejected": rejected_text}
|
362 |
return {"prompt": prompt_val, "chosen": chosen_val, "rejected": rejected_val}
|
363 |
|
364 |
+
dataset = dataset.map(to_pref, remove_columns=dataset.column_names, num_proc=num_proc)
|
365 |
return dataset
|
366 |
|
367 |
# If we reach here, we don't have required fields for DPO
|
|
|
391 |
"output": completion
|
392 |
}
|
393 |
|
394 |
+
dataset = dataset.map(format_openhermes_fr, remove_columns=dataset.column_names, num_proc=num_proc)
|
395 |
|
396 |
elif dataset_format == "messages":
|
397 |
# Process messages format (like HuggingFaceH4/Multilingual-Thinking)
|
|
|
436 |
|
437 |
return {"text": text}
|
438 |
|
439 |
+
dataset = dataset.map(format_messages, remove_columns=dataset.column_names, num_proc=num_proc)
|
440 |
|
441 |
elif dataset_format == "text":
|
442 |
# Process plain text format
|
|
|
447 |
text += "</s>"
|
448 |
return {"text": text}
|
449 |
|
450 |
+
dataset = dataset.map(format_text, remove_columns=dataset.column_names, num_proc=num_proc)
|
451 |
|
452 |
elif dataset_format == "custom":
|
453 |
# Custom format - user handles this in their config
|
|
|
672 |
"bf16": bf16,
|
673 |
# Some versions support tf32
|
674 |
"tf32": tf32 if 'tf32' in TrainingArguments.__init__.__code__.co_varnames else None,
|
675 |
+
# Optimizer (optionally use fused AdamW if available through config)
|
676 |
+
"optim": getattr(config, 'optimizer', 'adamw_torch'),
|
677 |
# Regularization
|
678 |
"weight_decay": weight_decay,
|
679 |
"max_grad_norm": max_grad_norm,
|
|
|
850 |
if "max_seq_length" in sft_params:
|
851 |
sft_kwargs["max_seq_length"] = getattr(config, 'max_seq_length', 2048)
|
852 |
|
853 |
+
# Enable sequence packing if supported by TRL (speeds up token utilization)
|
854 |
+
if "packing" in sft_params:
|
855 |
+
sft_kwargs["packing"] = getattr(config, 'packing', False)
|
856 |
+
|
857 |
# Remove any None values
|
858 |
sft_kwargs = {k: v for k, v in sft_kwargs.items() if v is not None}
|
859 |
|
src/dataset_utils.py
CHANGED
@@ -122,12 +122,20 @@ class TrackioDatasetManager:
|
|
122 |
|
123 |
def save_experiments(self, experiments: List[Dict[str, Any]], commit_message: Optional[str] = None) -> bool:
|
124 |
"""
|
125 |
-
Save a list of experiments to the dataset
|
126 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
127 |
Args:
|
128 |
experiments (List[Dict[str, Any]]): List of experiment dictionaries
|
129 |
commit_message (Optional[str]): Custom commit message
|
130 |
-
|
131 |
Returns:
|
132 |
bool: True if save was successful, False otherwise
|
133 |
"""
|
@@ -136,24 +144,120 @@ class TrackioDatasetManager:
|
|
136 |
logger.warning("⚠️ No experiments to save")
|
137 |
return False
|
138 |
|
139 |
-
#
|
140 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
141 |
for exp in experiments:
|
142 |
-
if self._validate_experiment_structure(exp):
|
143 |
-
# Ensure last_updated is set
|
144 |
-
if 'last_updated' not in exp:
|
145 |
-
exp['last_updated'] = datetime.now().isoformat()
|
146 |
-
valid_experiments.append(exp)
|
147 |
-
else:
|
148 |
logger.error(f"❌ Invalid experiment structure: {exp.get('experiment_id', 'unknown')}")
|
149 |
return False
|
150 |
-
|
151 |
-
|
152 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
|
154 |
# Generate commit message if not provided
|
155 |
if not commit_message:
|
156 |
-
commit_message = f"
|
157 |
|
158 |
# Push to hub
|
159 |
dataset.push_to_hub(
|
@@ -163,7 +267,7 @@ class TrackioDatasetManager:
|
|
163 |
commit_message=commit_message
|
164 |
)
|
165 |
|
166 |
-
logger.info(f"✅ Successfully saved {len(
|
167 |
return True
|
168 |
|
169 |
except Exception as e:
|
|
|
122 |
|
123 |
def save_experiments(self, experiments: List[Dict[str, Any]], commit_message: Optional[str] = None) -> bool:
|
124 |
"""
|
125 |
+
Save a list of experiments to the dataset using a non-destructive union merge.
|
126 |
+
|
127 |
+
- Loads existing experiments (if any) and builds a union by `experiment_id`.
|
128 |
+
- For overlapping IDs, merges JSON fields:
|
129 |
+
- metrics: concatenates lists and de-duplicates by (step, timestamp) for nested entries
|
130 |
+
- parameters: dict-update (new values override)
|
131 |
+
- artifacts: union with de-dup
|
132 |
+
- logs: concatenation with de-dup
|
133 |
+
- Non-JSON scalar fields from incoming experiments take precedence.
|
134 |
+
|
135 |
Args:
|
136 |
experiments (List[Dict[str, Any]]): List of experiment dictionaries
|
137 |
commit_message (Optional[str]): Custom commit message
|
138 |
+
|
139 |
Returns:
|
140 |
bool: True if save was successful, False otherwise
|
141 |
"""
|
|
|
144 |
logger.warning("⚠️ No experiments to save")
|
145 |
return False
|
146 |
|
147 |
+
# Helpers
|
148 |
+
def _parse_json_field(value, default):
|
149 |
+
try:
|
150 |
+
if value is None:
|
151 |
+
return default
|
152 |
+
if isinstance(value, str):
|
153 |
+
return json.loads(value) if value else default
|
154 |
+
return value
|
155 |
+
except Exception:
|
156 |
+
return default
|
157 |
+
|
158 |
+
def _metrics_key(entry: Dict[str, Any]):
|
159 |
+
if isinstance(entry, dict):
|
160 |
+
return (entry.get('step'), entry.get('timestamp'))
|
161 |
+
return (None, json.dumps(entry, sort_keys=True))
|
162 |
+
|
163 |
+
# Load existing experiments for union merge
|
164 |
+
existing = {}
|
165 |
+
try:
|
166 |
+
for row in self.load_existing_experiments():
|
167 |
+
exp_id = row.get('experiment_id')
|
168 |
+
if exp_id:
|
169 |
+
existing[exp_id] = row
|
170 |
+
except Exception:
|
171 |
+
existing = {}
|
172 |
+
|
173 |
+
# Validate and merge
|
174 |
+
merged_map: Dict[str, Dict[str, Any]] = {}
|
175 |
+
# Seed with existing
|
176 |
+
for exp_id, row in existing.items():
|
177 |
+
merged_map[exp_id] = row
|
178 |
+
|
179 |
+
# Apply incoming
|
180 |
for exp in experiments:
|
181 |
+
if not self._validate_experiment_structure(exp):
|
|
|
|
|
|
|
|
|
|
|
182 |
logger.error(f"❌ Invalid experiment structure: {exp.get('experiment_id', 'unknown')}")
|
183 |
return False
|
184 |
+
exp_id = exp['experiment_id']
|
185 |
+
incoming = exp
|
186 |
+
if exp_id not in merged_map:
|
187 |
+
incoming['last_updated'] = incoming.get('last_updated') or datetime.now().isoformat()
|
188 |
+
merged_map[exp_id] = incoming
|
189 |
+
continue
|
190 |
+
# Merge with existing
|
191 |
+
base = merged_map[exp_id]
|
192 |
+
# Parse JSON fields
|
193 |
+
base_metrics = _parse_json_field(base.get('metrics'), [])
|
194 |
+
base_params = _parse_json_field(base.get('parameters'), {})
|
195 |
+
base_artifacts = _parse_json_field(base.get('artifacts'), [])
|
196 |
+
base_logs = _parse_json_field(base.get('logs'), [])
|
197 |
+
inc_metrics = _parse_json_field(incoming.get('metrics'), [])
|
198 |
+
inc_params = _parse_json_field(incoming.get('parameters'), {})
|
199 |
+
inc_artifacts = _parse_json_field(incoming.get('artifacts'), [])
|
200 |
+
inc_logs = _parse_json_field(incoming.get('logs'), [])
|
201 |
+
# Merge metrics with de-dup
|
202 |
+
merged_metrics = []
|
203 |
+
seen = set()
|
204 |
+
for entry in base_metrics + inc_metrics:
|
205 |
+
try:
|
206 |
+
# Use the original entry so _metrics_key can properly
|
207 |
+
# distinguish dict vs non-dict entries
|
208 |
+
key = _metrics_key(entry)
|
209 |
+
except Exception:
|
210 |
+
key = (None, None)
|
211 |
+
if key not in seen:
|
212 |
+
seen.add(key)
|
213 |
+
merged_metrics.append(entry)
|
214 |
+
# Merge params
|
215 |
+
merged_params = {}
|
216 |
+
if isinstance(base_params, dict):
|
217 |
+
merged_params.update(base_params)
|
218 |
+
if isinstance(inc_params, dict):
|
219 |
+
merged_params.update(inc_params)
|
220 |
+
# Merge artifacts and logs with de-dup
|
221 |
+
def _dedup_list(lst):
|
222 |
+
out = []
|
223 |
+
seen_local = set()
|
224 |
+
for item in lst:
|
225 |
+
key = json.dumps(item, sort_keys=True, default=str) if not isinstance(item, str) else item
|
226 |
+
if key not in seen_local:
|
227 |
+
seen_local.add(key)
|
228 |
+
out.append(item)
|
229 |
+
return out
|
230 |
+
merged_artifacts = _dedup_list(list(base_artifacts) + list(inc_artifacts))
|
231 |
+
merged_logs = _dedup_list(list(base_logs) + list(inc_logs))
|
232 |
+
# Rebuild merged record preferring incoming scalars
|
233 |
+
merged_rec = dict(base)
|
234 |
+
merged_rec.update({k: v for k, v in incoming.items() if k not in ('metrics', 'parameters', 'artifacts', 'logs')})
|
235 |
+
merged_rec['metrics'] = json.dumps(merged_metrics, default=str)
|
236 |
+
merged_rec['parameters'] = json.dumps(merged_params, default=str)
|
237 |
+
merged_rec['artifacts'] = json.dumps(merged_artifacts, default=str)
|
238 |
+
merged_rec['logs'] = json.dumps(merged_logs, default=str)
|
239 |
+
merged_rec['last_updated'] = datetime.now().isoformat()
|
240 |
+
merged_map[exp_id] = merged_rec
|
241 |
+
|
242 |
+
# Prepare final list
|
243 |
+
valid_experiments = list(merged_map.values())
|
244 |
+
# Ensure all have mandatory fields encoded
|
245 |
+
normalized = []
|
246 |
+
for rec in valid_experiments:
|
247 |
+
# Normalize json fields to strings
|
248 |
+
for f, default in (('metrics', []), ('parameters', {}), ('artifacts', []), ('logs', [])):
|
249 |
+
val = rec.get(f)
|
250 |
+
if not isinstance(val, str):
|
251 |
+
rec[f] = json.dumps(val if val is not None else default, default=str)
|
252 |
+
if 'last_updated' not in rec:
|
253 |
+
rec['last_updated'] = datetime.now().isoformat()
|
254 |
+
normalized.append(rec)
|
255 |
+
|
256 |
+
dataset = Dataset.from_list(normalized)
|
257 |
|
258 |
# Generate commit message if not provided
|
259 |
if not commit_message:
|
260 |
+
commit_message = f"Union-merge update with {len(normalized)} experiments ({datetime.now().isoformat()})"
|
261 |
|
262 |
# Push to hub
|
263 |
dataset.push_to_hub(
|
|
|
267 |
commit_message=commit_message
|
268 |
)
|
269 |
|
270 |
+
logger.info(f"✅ Successfully saved {len(normalized)} experiments (union-merged) to {self.dataset_repo}")
|
271 |
return True
|
272 |
|
273 |
except Exception as e:
|