Bev / evaluations /published-examples.json
Reza2kn's picture
Package Bev v0.1.1 source, wheel, licenses and reproducible evaluations
71be121 verified
Raw
History Blame Contribute Delete
10.4 kB
{
"kind": "documentation example checks, not benchmark",
"passed": true,
"outputs": {
"support-request.json": {
"model": "bev-bonsai-27b",
"backend": "prism_llama_cpp",
"backend_build": "b1-9a9394a",
"mode": "llama_full_vocab_categorical_scoring",
"strategy": "batch",
"requested_strategy": "auto",
"strategy_implementation": "bounded_concurrent_native_requests",
"parsed_json": {
"route": "billing"
},
"fields": {
"route": {
"value": "billing",
"selected_value": "billing",
"selected_label": "A",
"selected_token_id": 32,
"probability": 0.9989061260210375,
"abstained": false,
"candidate_probability_mass": 0.9086667884355636,
"candidates": [
{
"value": "billing",
"label": "A",
"token_id": 32,
"probability": 0.9989061260210375,
"logprob": -0.09687129408121109
},
{
"value": "technical",
"label": "B",
"token_id": 33,
"probability": 0.0008516050271616897,
"logprob": -7.1641645431518555
},
{
"value": "unknown",
"label": "C",
"token_id": 34,
"probability": 0.0002422689518008115,
"logprob": -8.421238899230957
}
],
"elapsed_ms": 474.855,
"queue_ms": 0.003,
"backend_requests": 1,
"score_retrieval": "selected_token_logprobs",
"requested_top_n": 3,
"prompt_tokens": 142,
"backend_timings": {
"cache_n": 0,
"prompt_n": 142,
"prompt_ms": 284.624,
"prompt_per_token_ms": 2.0043943661971833,
"prompt_per_second": 498.90381696554044,
"predicted_n": 1,
"predicted_ms": 0.001,
"predicted_per_token_ms": 0.0,
"predicted_per_second": 0.0
},
"usage": {
"prompt_tokens": 142,
"completion_tokens": 1,
"total_tokens": 143
}
}
},
"scores_are_calibrated": false,
"score_source": "pre_sampling_full_vocabulary_logprobs",
"selected_token_logprobs": true,
"score_temperature": 1.0,
"abstained_fields": [],
"elapsed_ms": 480.079,
"prompt_preparation_ms": 5.143,
"backend_requests": 1,
"scored_fields": 1,
"max_prompt_tokens": 142,
"max_model_len": 16384,
"backend_slots": 2,
"cache_block_tokens": null,
"usage": {
"prompt_tokens": 142,
"completion_tokens": 1,
"total_tokens": 143
}
},
"systemone-request.json": {
"model": "bev-bonsai-27b",
"answers": {
"queue": {
"type": "choice",
"choice": "billing",
"probabilities": {
"billing": 0.9982291183228973,
"technical": 0.0016640844610198414,
"general": 0.00010679721608279538
},
"confidence": 0.9982291183228973
},
"outage": {
"type": "noul",
"noul": 0.002787811663481365
},
"priority": {
"type": "score",
"score": 1.0002383132357986,
"probabilities": {
"0": 0.0007058337315065528,
"1": 0.9983500193011885,
"2": 0.0009441469673050024
},
"legend": {
"0": "General question",
"1": "Billing issue",
"2": "Outage"
},
"confidence": 0.9983500193011885
}
},
"usage": {
"input_tokens": 547,
"output_tokens": 3,
"prompt_tokens": 547,
"completion_tokens": 3,
"total_tokens": 550
},
"usage_accounting": "logical_prompt_and_scoring_tokens_including_retries",
"scores_are_calibrated": false,
"confidence_definition": "maximum_candidate_relative_probability",
"evaluation_trace": {
"model": "bev-bonsai-27b",
"backend": "prism_llama_cpp",
"backend_build": "b1-9a9394a",
"mode": "llama_full_vocab_categorical_scoring",
"strategy": "batch",
"requested_strategy": "auto",
"strategy_implementation": "bounded_concurrent_native_requests",
"parsed_json": {
"queue": "billing",
"outage": false,
"priority": "1"
},
"fields": {
"queue": {
"value": "billing",
"selected_value": "billing",
"selected_label": "A",
"selected_token_id": 32,
"probability": 0.9982291183228973,
"abstained": false,
"candidate_probability_mass": 0.8655803643819256,
"candidates": [
{
"value": "billing",
"label": "A",
"token_id": 32,
"probability": 0.9982291183228973,
"logprob": -0.14612750709056854
},
{
"value": "technical",
"label": "B",
"token_id": 33,
"probability": 0.0016640844610198414,
"logprob": -6.542835235595703
},
{
"value": "general",
"label": "C",
"token_id": 34,
"probability": 0.00010679721608279538,
"logprob": -9.288933753967285
}
],
"elapsed_ms": 576.291,
"queue_ms": 0.002,
"backend_requests": 1,
"score_retrieval": "selected_token_logprobs",
"requested_top_n": 3,
"prompt_tokens": 190,
"backend_timings": {
"cache_n": 0,
"prompt_n": 190,
"prompt_ms": 572.964,
"prompt_per_token_ms": 3.0156000000000005,
"prompt_per_second": 331.60896670645974,
"predicted_n": 1,
"predicted_ms": 0.001,
"predicted_per_token_ms": 0.0,
"predicted_per_second": 0.0
},
"usage": {
"prompt_tokens": 190,
"completion_tokens": 1,
"total_tokens": 191
}
},
"outage": {
"value": false,
"selected_value": false,
"selected_label": "B",
"selected_token_id": 33,
"probability": 0.9972121883365187,
"abstained": false,
"candidate_probability_mass": 0.9862726128870609,
"candidates": [
{
"value": true,
"label": "A",
"token_id": 32,
"probability": 0.002787811663481365,
"logprob": -5.896320819854736
},
{
"value": false,
"label": "B",
"token_id": 33,
"probability": 0.9972121883365187,
"logprob": -0.01661418378353119
}
],
"elapsed_ms": 578.692,
"queue_ms": 0.001,
"backend_requests": 1,
"score_retrieval": "selected_token_logprobs",
"requested_top_n": 2,
"prompt_tokens": 164,
"backend_timings": {
"cache_n": 0,
"prompt_n": 164,
"prompt_ms": 570.737,
"prompt_per_token_ms": 3.480103658536585,
"prompt_per_second": 287.34776262972264,
"predicted_n": 1,
"predicted_ms": 0.001,
"predicted_per_token_ms": 0.0,
"predicted_per_second": 0.0
},
"usage": {
"prompt_tokens": 164,
"completion_tokens": 1,
"total_tokens": 165
}
},
"priority": {
"value": "1",
"selected_value": "1",
"selected_label": "B",
"selected_token_id": 33,
"probability": 0.9983500193011885,
"abstained": false,
"candidate_probability_mass": 0.9648334149708054,
"candidates": [
{
"value": "0",
"label": "A",
"token_id": 32,
"probability": 0.0007058337315065528,
"logprob": -7.291930675506592
},
{
"value": "1",
"label": "B",
"token_id": 33,
"probability": 0.9983500193011885,
"logprob": -0.037451162934303284
},
{
"value": "2",
"label": "C",
"token_id": 34,
"probability": 0.0009441469673050024,
"logprob": -7.001028537750244
}
],
"elapsed_ms": 977.959,
"queue_ms": 576.01,
"backend_requests": 1,
"score_retrieval": "selected_token_logprobs",
"requested_top_n": 3,
"prompt_tokens": 193,
"backend_timings": {
"cache_n": 0,
"prompt_n": 193,
"prompt_ms": 314.069,
"prompt_per_token_ms": 1.627300518134715,
"prompt_per_second": 614.5146448710315,
"predicted_n": 1,
"predicted_ms": 0.001,
"predicted_per_token_ms": 0.0,
"predicted_per_second": 0.0
},
"usage": {
"prompt_tokens": 193,
"completion_tokens": 1,
"total_tokens": 194
}
}
},
"scores_are_calibrated": false,
"score_source": "pre_sampling_full_vocabulary_logprobs",
"selected_token_logprobs": true,
"score_temperature": 1.0,
"abstained_fields": [],
"elapsed_ms": 983.191,
"prompt_preparation_ms": 4.88,
"backend_requests": 3,
"scored_fields": 3,
"max_prompt_tokens": 193,
"max_model_len": 16384,
"backend_slots": 2,
"cache_block_tokens": null,
"usage": {
"prompt_tokens": 547,
"completion_tokens": 3,
"total_tokens": 550
}
}
}
}
}