{
"model_details": {
"provider": "e-infra",
"model_requested": "deepseek-v3.2-thinking",
"model_for_requests": "deepseek-v3.2-thinking",
"api_base_url": "https://llm.ai.e-infra.cz/v1",
"chat_completions_endpoint": "https://llm.ai.e-infra.cz/v1/chat/completions"
},
"run_config": {
"input": [
"data/input/prepositions.csv"
],
"labels": "",
"task_name": "Err. correct. prepositions",
"task_description": "Correcting errors in prepositions in a transcription of learners' dialogue.",
"tags": "error;correction;preposition;English",
"model": "deepseek-v3.2-thinking",
"temperature": null,
"top_p": null,
"top_k": null,
"service_tier": "standard",
"verbosity": null,
"reasoning_effort": null,
"thinking_level": null,
"effort": null,
"strict_control_acceptance": true,
"provider": "e-infra",
"system_prompt": "You are a meticulous annotator focusing on grammatical errors in the use of prepositions.\nYou will receive a short excerpt of a transcribed dialogue where the node word is a preposition.\nDecide whether it is a correct preposition given the context.\n\nReturn the correct preposition (even if it already is correct). \nIf the preposition should be omitted (i.e. the grammatical correct solution is leaving it out), return 0.\n\nIn your decisions, follow these rules:\n- Do not correct unless necessary: if the preposition is acceptable, even if non-standard (but used in a variety of English) or less idiomatic. In such cases return the preposition as is.\n- Expect spoken language features such as fragments, disfluencies, repetitions, and false starts; treat transcription artifacts (e.g. , , , pauses, broken words) as signals of turn structure rather than reliable syntax. If the preposition appears within a short turn delimited by and , prioritize interpreting it within that turn rather than relying on the broader context. \n- If a correct preposition is repeated, which usually signals hesitation, do not correct it. In such cases return the preposition as is. \n- If the speaker self-corrects, evaluate the final intended form.\n- Expect other errors in tense, word choice, or sentence structure, and if the issue is not the preposition itself, do not correct it. In such cases return the preposition as is.\n- If multiple prepositions are possible and none is clearly wrong, do not correct it. In such cases return the preposition as is.",
"system_prompt_b64": null,
"few_shot_examples": 0,
"prompt_layout": "standard",
"cache_pad_target_tokens": 0,
"prompt_cache_key": null,
"gemini_cached_content": null,
"requesty_auto_cache": null,
"vertex_auto_adc_login": null,
"vertex_access_token_refresh_seconds": null,
"create_gemini_cache": false,
"gemini_cache_ttl": 3600,
"gemini_cache_ttl_autoupdate": true,
"keep_gemini_cache": false,
"enable_cot": true,
"no_explanation": false,
"logprobs": true,
"calibration": true,
"confusion_heatmap": true,
"api_key_var": "E-INFRA_API_KEY",
"api_base_var": "E-INFRA_BASE_URL",
"max_retries": 3,
"retry_delay": 5.0,
"request_interval_ms": 0,
"request_timeout_seconds": 600.0,
"threads": 3,
"prompt_log_detail": "full",
"flush_rows": 100,
"flush_seconds": 2.0,
"validator_cmd": null,
"validator_args": "",
"validator_timeout": 5.0,
"validator_prompt_max_candidates": 50,
"validator_prompt_max_chars": 8000,
"validator_exhausted_policy": "accept_blank_confidence",
"validator_debug": false,
"log_level": "INFO"
},
"source_input_csv": "data/input/prepositions.csv",
"source_output_csv": "data/output/prepositions__einfra__deepseekv32thinking__2026-04-02-13-14.csv",
"source_labels_csv": "",
"cache_padding": {
"enabled": false,
"target_shared_prefix_tokens": 0,
"calibration_shared_prefix_tokens": null,
"target_prompt_tokens": 0,
"calibration_prompt_tokens": null,
"calibration_example_id": null,
"applied_padding_tokens_estimate": 0,
"examples_with_padding_applied": 0
},
"request_control_summary": {
"configured": {},
"attempts_total": 1283,
"attempts_with_control_telemetry": 1283,
"per_control": {
"reasoning_effort": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
},
"thinking_level": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
},
"effort": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
},
"verbosity": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
},
"prompt_cache_key": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
},
"gemini_cached_content": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
},
"requesty_auto_cache": {
"configured_value": null,
"requested_attempts": 0,
"sent_attempts": 0,
"accepted_attempts": 0,
"rejected_attempts": 0,
"missing_from_final_request_attempts": 0,
"acceptance_rate": null,
"rejected_reasons": {},
"rejected_example_ids": []
}
}
},
"usage_metadata_summary": {
"attempts_total": 1283,
"attempts_with_usage_metadata": 1283,
"attempts_with_cached_token_signals": 0,
"cached_tokens_total_estimate": 0,
"cache_read_tokens_total": 0,
"cache_write_tokens_total": 0,
"cache_token_fields_totals": {},
"attempts_with_gemini_cached_content_token_signals": 0,
"gemini_cached_content_token_count_total": 0,
"gemini_cached_content_token_fields_totals": {}
},
"token_usage_totals": {
"attempts_total": 1283,
"attempts_with_token_usage": 1283,
"attempts_with_output_tokens": 1283,
"attempts_with_cached_input_tokens": 0,
"attempts_with_thinking_tokens": 0,
"input_tokens_total": 1065302,
"cached_input_tokens_total": 0,
"non_cached_input_tokens_total": 1065302,
"output_tokens_total": 1707247,
"thinking_tokens_total": 0,
"output_tokens_definition": "total_tokens - prompt_tokens (or completion_tokens + thinking_tokens fallback)"
},
"truth_label_count": 1283,
"prediction_count": 1283,
"evaluated_example_count": 1283,
"calibration_metrics": {
"available": true,
"sample_count": 1283,
"bin_count": 10,
"ece": 0.027233047544820567,
"mce": 0.1668674698795165,
"brier_score": 0.0832159002338269
},
"first_prompt_timestamp": "2026-04-02T11:14:24.260967Z",
"last_prompt_timestamp": "2026-04-02T14:01:04.329339Z",
"overall_time_seconds": 10000.068372,
"overall_time_human": "2h 46m 40s",
"accuracy": 0.8924395946999221,
"cohen_kappa": 0.7595338824791694,
"macro_precision": 0.3994867040030304,
"macro_recall": 0.3878165811941068,
"macro_f1": 0.37918792404179286,
"per_label": {
"0": {
"precision": 0.3076923076923077,
"recall": 0.48,
"f1": 0.37500000000000006,
"support": 25
},
"about": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 2
},
"around": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 1
},
"as": {
"precision": 1.0,
"recall": 1.0,
"f1": 1.0,
"support": 1
},
"at": {
"precision": 0.9263157894736842,
"recall": 0.8756218905472637,
"f1": 0.9002557544757034,
"support": 201
},
"behind": {
"precision": 1.0,
"recall": 1.0,
"f1": 1.0,
"support": 1
},
"during": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 4
},
"for": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 0
},
"from": {
"precision": 0.3333333333333333,
"recall": 0.5,
"f1": 0.4,
"support": 2
},
"in": {
"precision": 0.928495197438634,
"recall": 0.9466811751904244,
"f1": 0.9375,
"support": 919
},
"into": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 1
},
"of": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 2
},
"on": {
"precision": 0.7821782178217822,
"recall": 0.8586956521739131,
"f1": 0.8186528497409327,
"support": 92
},
"to": {
"precision": 0.7142857142857143,
"recall": 0.15625,
"f1": 0.25641025641025644,
"support": 32
},
"with": {
"precision": 0.0,
"recall": 0.0,
"f1": 0.0,
"support": 0
}
},
"labels": [
"0",
"about",
"around",
"as",
"at",
"behind",
"during",
"for",
"from",
"in",
"into",
"of",
"on",
"to",
"with"
],
"label_count": 15,
"total_examples": 1283,
"confusion_matrix": {
"0": {
"0": 12,
"about": 0,
"around": 0,
"as": 0,
"at": 2,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 9,
"into": 0,
"of": 0,
"on": 2,
"to": 0,
"with": 0
},
"about": {
"0": 1,
"about": 0,
"around": 0,
"as": 0,
"at": 1,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 0,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"around": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 1,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"as": {
"0": 0,
"about": 0,
"around": 0,
"as": 1,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 0,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"at": {
"0": 1,
"about": 0,
"around": 0,
"as": 0,
"at": 176,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 16,
"into": 0,
"of": 0,
"on": 8,
"to": 0,
"with": 0
},
"behind": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 1,
"during": 0,
"for": 0,
"from": 0,
"in": 0,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"during": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 3,
"into": 0,
"of": 0,
"on": 1,
"to": 0,
"with": 0
},
"for": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 0,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"from": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 1,
"in": 1,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"in": {
"0": 23,
"about": 0,
"around": 0,
"as": 0,
"at": 10,
"behind": 0,
"during": 0,
"for": 1,
"from": 2,
"in": 870,
"into": 0,
"of": 1,
"on": 9,
"to": 2,
"with": 1
},
"into": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 1,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
},
"of": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 1,
"into": 0,
"of": 0,
"on": 1,
"to": 0,
"with": 0
},
"on": {
"0": 2,
"about": 0,
"around": 0,
"as": 0,
"at": 1,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 10,
"into": 0,
"of": 0,
"on": 79,
"to": 0,
"with": 0
},
"to": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 25,
"into": 0,
"of": 0,
"on": 1,
"to": 5,
"with": 1
},
"with": {
"0": 0,
"about": 0,
"around": 0,
"as": 0,
"at": 0,
"behind": 0,
"during": 0,
"for": 0,
"from": 0,
"in": 0,
"into": 0,
"of": 0,
"on": 0,
"to": 0,
"with": 0
}
},
"label_metrics_available": true
}