developer_uid: rirv938
submission_id: chaiml-llama-8b-multih_78780_v29
model_name: chaiml-llama-8b-multih_78780_v29
model_group: ChaiML/llama_8b_multihea
status: torndown
timestamp: 2025-02-14T01:57:38+00:00
num_battles: 5610
num_wins: 2968
celo_rating: 1280.78
family_friendly_score: 0.5602
family_friendly_standard_error: 0.007019629050028214
submission_type: basic
model_repo: ChaiML/llama_8b_multihead_204m_512_v3_tokens_step_398208
model_architecture: MultiHeadLlamaClassifier
model_num_parameters: 8030261248.0
best_of: 1
max_input_tokens: 256
max_output_tokens: 1
display_name: chaiml-llama-8b-multih_78780_v29
ineligible_reason: max_output_tokens!=64
is_internal_developer: True
language_model: ChaiML/llama_8b_multihead_204m_512_v3_tokens_step_398208
model_size: 8B
ranking_group: single
us_pacific_date: 2025-02-13
win_ratio: 0.5290552584670232
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 256, 'best_of': 1, 'max_output_tokens': 1}
formatter: {'memory_template': '', 'prompt_template': '', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '', 'truncate_by_message': False}
Resubmit model
Shutdown handler not registered because Python interpreter is not running in the main thread
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLizer
Starting job with name chaiml-llama-8b-multih-78780-v29-mkmlizer
Waiting for job on chaiml-llama-8b-multih-78780-v29-mkmlizer to finish
chaiml-llama-8b-multih-78780-v29-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║     _____            __           __                                ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║       /___/                                                         ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║                                                                     ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  Version: 0.12.8                                                    ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║                                                                     ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  belonging to:                                                      ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║                                                                     ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║  Expiration: 2025-04-15 23:59:59                                    ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ║                                                                     ║
chaiml-llama-8b-multih-78780-v29-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-llama-8b-multih-78780-v29-mkmlizer: Downloaded to shared memory in 23.339s
chaiml-llama-8b-multih-78780-v29-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpv0khyp84, device:0
chaiml-llama-8b-multih-78780-v29-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-llama-8b-multih-78780-v29-mkmlizer: quantized model in 15.773s
chaiml-llama-8b-multih-78780-v29-mkmlizer: Processed model ChaiML/llama_8b_multihead_204m_512_v3_tokens_step_398208 in 39.112s
chaiml-llama-8b-multih-78780-v29-mkmlizer: creating bucket guanaco-mkml-models
chaiml-llama-8b-multih-78780-v29-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-llama-8b-multih-78780-v29-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29
chaiml-llama-8b-multih-78780-v29-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29/config.json
chaiml-llama-8b-multih-78780-v29-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29/special_tokens_map.json
chaiml-llama-8b-multih-78780-v29-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29/tokenizer_config.json
chaiml-llama-8b-multih-78780-v29-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29/tokenizer.json
chaiml-llama-8b-multih-78780-v29-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29/flywheel_model.0.safetensors
chaiml-llama-8b-multih-78780-v29-mkmlizer: 
Loading 0:   0%|          | 0/294 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 5/294 [00:00<00:08, 32.34it/s]
Loading 0:   4%|▍         | 13/294 [00:00<00:05, 51.65it/s]
Loading 0:   6%|▋         | 19/294 [00:00<00:05, 47.61it/s]
Loading 0:   8%|▊         | 24/294 [00:00<00:05, 47.03it/s]
Loading 0:  11%|█         | 31/294 [00:00<00:05, 52.17it/s]
Loading 0:  13%|█▎        | 37/294 [00:00<00:05, 50.17it/s]
Loading 0:  15%|█▍        | 43/294 [00:00<00:04, 52.73it/s]
Loading 0:  17%|█▋        | 50/294 [00:01<00:04, 49.45it/s]
Loading 0:  20%|█▉        | 58/294 [00:01<00:04, 56.62it/s]
Loading 0:  22%|██▏       | 64/294 [00:01<00:04, 52.78it/s]
Loading 0:  24%|██▍       | 70/294 [00:01<00:04, 53.23it/s]
Loading 0:  26%|██▌       | 76/294 [00:01<00:04, 54.20it/s]
Loading 0:  28%|██▊       | 82/294 [00:01<00:04, 51.91it/s]
Loading 0:  30%|██▉       | 88/294 [00:01<00:05, 38.39it/s]
Loading 0:  32%|███▏      | 95/294 [00:02<00:05, 39.79it/s]
Loading 0:  35%|███▍      | 102/294 [00:02<00:04, 45.91it/s]
Loading 0:  37%|███▋      | 108/294 [00:02<00:03, 49.06it/s]
Loading 0:  39%|███▉      | 114/294 [00:02<00:03, 46.76it/s]
Loading 0:  41%|████▏     | 122/294 [00:02<00:03, 47.65it/s]
Loading 0:  44%|████▍     | 130/294 [00:02<00:02, 55.01it/s]
Loading 0:  46%|████▋     | 136/294 [00:02<00:03, 50.42it/s]
Loading 0:  48%|████▊     | 142/294 [00:02<00:02, 52.02it/s]
Loading 0:  51%|█████     | 149/294 [00:03<00:02, 49.54it/s]
Loading 0:  53%|█████▎    | 157/294 [00:03<00:02, 56.03it/s]
Loading 0:  55%|█████▌    | 163/294 [00:03<00:02, 51.89it/s]
Loading 0:  57%|█████▋    | 169/294 [00:03<00:02, 53.72it/s]
Loading 0:  60%|██████    | 177/294 [00:03<00:02, 56.04it/s]
Loading 0:  62%|██████▏   | 183/294 [00:03<00:01, 56.37it/s]
Loading 0:  64%|██████▍   | 189/294 [00:03<00:02, 38.76it/s]
Loading 0:  66%|██████▌   | 194/294 [00:03<00:02, 40.96it/s]
Loading 0:  69%|██████▊   | 202/294 [00:04<00:01, 49.65it/s]
Loading 0:  71%|███████   | 208/294 [00:04<00:01, 48.90it/s]
Loading 0:  73%|███████▎  | 214/294 [00:04<00:01, 48.24it/s]
Loading 0:  75%|███████▍  | 220/294 [00:04<00:01, 50.71it/s]
Loading 0:  77%|███████▋  | 226/294 [00:04<00:01, 49.74it/s]
Loading 0:  79%|███████▉  | 232/294 [00:04<00:01, 52.26it/s]
Loading 0:  81%|████████▏ | 239/294 [00:04<00:01, 49.29it/s]
Loading 0:  84%|████████▍ | 248/294 [00:05<00:00, 52.13it/s]
Loading 0:  87%|████████▋ | 257/294 [00:05<00:00, 52.56it/s]
Loading 0:  90%|█████████ | 265/294 [00:05<00:00, 56.99it/s]
Loading 0:  92%|█████████▏| 271/294 [00:05<00:00, 54.55it/s]
Loading 0:  94%|█████████▍| 277/294 [00:05<00:00, 55.73it/s]
Loading 0:  96%|█████████▋| 283/294 [00:05<00:00, 51.27it/s]
Loading 0:  98%|█████████▊| 289/294 [00:05<00:00, 41.97it/s]
                                                            

Job chaiml-llama-8b-multih-78780-v29-mkmlizer completed after 205.68s with status: succeeded
Stopping job with name chaiml-llama-8b-multih-78780-v29-mkmlizer
Pipeline stage MKMLizer completed in 206.43s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.16s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service chaiml-llama-8b-multih-78780-v29
Waiting for inference service chaiml-llama-8b-multih-78780-v29 to be ready
Inference service chaiml-llama-8b-multih-78780-v29 ready after 180.65786027908325s
Pipeline stage MKMLDeployer completed in 181.19s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.9607582092285156s
Received healthy response to inference request in 5.768481731414795s
Received healthy response to inference request in 2.2068381309509277s
Received healthy response to inference request in 3.2173736095428467s
Received healthy response to inference request in 4.099485158920288s
5 requests
0 failed requests
5th percentile: 2.0099741935729982
10th percentile: 2.0591901779174804
20th percentile: 2.157622146606445
30th percentile: 2.4089452266693114
40th percentile: 2.8131594181060793
50th percentile: 3.2173736095428467
60th percentile: 3.570218229293823
70th percentile: 3.9230628490447996
80th percentile: 4.43328447341919
90th percentile: 5.100883102416993
95th percentile: 5.434682416915893
99th percentile: 5.701721868515015
mean time: 3.4505873680114747
%s, retrying in %s seconds...
Received healthy response to inference request in 3.1681742668151855s
Received healthy response to inference request in 3.355217218399048s
Received healthy response to inference request in 2.59489369392395s
Received healthy response to inference request in 2.590437173843384s
Received healthy response to inference request in 3.8579936027526855s
5 requests
0 failed requests
5th percentile: 2.591328477859497
10th percentile: 2.5922197818756105
20th percentile: 2.594002389907837
30th percentile: 2.709549808502197
40th percentile: 2.9388620376586916
50th percentile: 3.1681742668151855
60th percentile: 3.2429914474487305
70th percentile: 3.3178086280822754
80th percentile: 3.4557724952697755
90th percentile: 3.6568830490112303
95th percentile: 3.757438325881958
99th percentile: 3.83788254737854
mean time: 3.1133431911468508
Pipeline stage StressChecker completed in 35.78s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyTriggerPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage OfflineFamilyFriendlyTriggerPipeline completed in 0.71s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 0.67s
Shutdown handler de-registered
chaiml-llama-8b-multih_78780_v29 status is now deployed due to DeploymentManager action
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyScorer
Evaluating %s Family Friendly Score with %s threads
Retrying (%r) after connection broken by '%r': %s
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.14s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
%s, retrying in %s seconds...
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
%s, retrying in %s seconds...
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
clean up pipeline due to error=DeploymentError('Timeout to start the InferenceService chaiml-llama-8b-multih-78780-v29-profiler.                                The InferenceService is as following: {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'kind\': \'InferenceService\', \'metadata\': {\'annotations\': {\'autoscaling.knative.dev/class\': \'hpa.autoscaling.knative.dev\', \'autoscaling.knative.dev/container-concurrency-target-percentage\': \'70\', \'autoscaling.knative.dev/initial-scale\': \'1\', \'autoscaling.knative.dev/max-scale-down-rate\': \'1.1\', \'autoscaling.knative.dev/max-scale-up-rate\': \'2\', \'autoscaling.knative.dev/metric\': \'mean_pod_latency_ms_v2\', \'autoscaling.knative.dev/panic-threshold-percentage\': \'650\', \'autoscaling.knative.dev/panic-window-percentage\': \'35\', \'autoscaling.knative.dev/scale-down-delay\': \'30s\', \'autoscaling.knative.dev/scale-to-zero-grace-period\': \'10m\', \'autoscaling.knative.dev/stable-window\': \'180s\', \'autoscaling.knative.dev/target\': \'3700\', \'autoscaling.knative.dev/target-burst-capacity\': \'-1\', \'autoscaling.knative.dev/tick-interval\': \'15s\', \'features.knative.dev/http-full-duplex\': \'Enabled\', \'networking.knative.dev/ingress-class\': \'istio.ingress.networking.knative.dev\'}, \'creationTimestamp\': \'2025-02-14T02:26:30Z\', \'finalizers\': [\'inferenceservice.finalizers\'], \'generation\': 1, \'labels\': {\'knative.coreweave.cloud/ingress\': \'istio.ingress.networking.knative.dev\', \'prometheus.k.chaiverse.com\': \'true\', \'qos.coreweave.cloud/latency\': \'low\'}, \'managedFields\': [{\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:metadata\': {\'f:annotations\': {\'.\': {}, \'f:autoscaling.knative.dev/class\': {}, \'f:autoscaling.knative.dev/container-concurrency-target-percentage\': {}, \'f:autoscaling.knative.dev/initial-scale\': {}, \'f:autoscaling.knative.dev/max-scale-down-rate\': {}, \'f:autoscaling.knative.dev/max-scale-up-rate\': {}, \'f:autoscaling.knative.dev/metric\': {}, \'f:autoscaling.knative.dev/panic-threshold-percentage\': {}, \'f:autoscaling.knative.dev/panic-window-percentage\': {}, \'f:autoscaling.knative.dev/scale-down-delay\': {}, \'f:autoscaling.knative.dev/scale-to-zero-grace-period\': {}, \'f:autoscaling.knative.dev/stable-window\': {}, \'f:autoscaling.knative.dev/target\': {}, \'f:autoscaling.knative.dev/target-burst-capacity\': {}, \'f:autoscaling.knative.dev/tick-interval\': {}, \'f:features.knative.dev/http-full-duplex\': {}, \'f:networking.knative.dev/ingress-class\': {}}, \'f:labels\': {\'.\': {}, \'f:knative.coreweave.cloud/ingress\': {}, \'f:prometheus.k.chaiverse.com\': {}, \'f:qos.coreweave.cloud/latency\': {}}}, \'f:spec\': {\'.\': {}, \'f:predictor\': {\'.\': {}, \'f:affinity\': {\'.\': {}, \'f:nodeAffinity\': {\'.\': {}, \'f:tion\': {}, \'f:requiredDuringSchedulingIgnoredDuringExecution\': {}}}, \'f:containerConcurrency\': {}, \'f:containers\': {}, \'f:imagePullSecrets\': {}, \'f:maxReplicas\': {}, \'f:minReplicas\': {}, \'f:timeout\': {}, \'f:volumes\': {}}}}, \'manager\': \'OpenAPI-Generator\', \'operation\': \'Update\', \'time\': \'2025-02-14T02:26:30Z\'}, {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:metadata\': {\'f:finalizers\': {\'.\': {}, \'v:"inferenceservice.finalizers"\': {}}}}, \'manager\': \'manager\', \'operation\': \'Update\', \'time\': \'2025-02-14T02:26:30Z\'}, {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:status\': {\'.\': {}, \'f:components\': {\'.\': {}, \'f:predictor\': {\'.\': {}, \'f:latestCreatedRevision\': {}}}, \'f:conditions\': {}, \'f:modelStatus\': {\'.\': {}, \'f:lastFailureInfo\': {\'.\': {}, \'f:exitCode\': {}, \'f:message\': {}, \'f:reason\': {}}, \'f:states\': {\'.\': {}, \'f:activeModelState\': {}, \'f:targetModelState\': {}}, \'f:transitionStatus\': {}}, \'f:observedGeneration\': {}}}, \'manager\': \'manager\', \'operation\': \'Update\', \'subresource\': \'status\', \'time\': \'2025-02-14T02:28:09Z\'}], \'name\': \'chaiml-llama-8b-multih-78780-v29-profiler\', \'namespace\': \'tenant-chaiml-guanaco\', \'resourceVersion\': \'275888578\', \'uid\': \'4a06380b-e7c9-4dcf-842e-e56cce65412f\'}, \'spec\': {\'predictor\': {\'affinity\': {\'nodeAffinity\': {\'tion\': [{\'preference\': {\'matchExpressions\': [{\'key\': \'topology.kubernetes.io/region\', \'operator\': \'In\', \'values\': [\'ORD1\']}]}, \'weight\': 5}], \'requiredDuringSchedulingIgnoredDuringExecution\': {\'nodeSelectorTerms\': [{\'matchExpressions\': [{\'key\': \'gpu.nvidia.com/class\', \'operator\': \'In\', \'values\': [\'RTX_A5000\']}]}]}}}, \'containerConcurrency\': 0, \'containers\': [{\'env\': [{\'name\': \'MAX_TOKEN_INPUT\', \'value\': \'256\'}, {\'name\': \'BEST_OF\', \'value\': \'1\'}, {\'name\': \'TEMPERATURE\', \'value\': \'1.0\'}, {\'name\': \'PRESENCE_PENALTY\', \'value\': \'0.0\'}, {\'name\': \'FREQUENCY_PENALTY\', \'value\': \'0.0\'}, {\'name\': \'TOP_P\', \'value\': \'1.0\'}, {\'name\': \'MIN_P\', \'value\': \'0.0\'}, {\'name\': \'TOP_K\', \'value\': \'40\'}, {\'name\': \'STOPPING_WORDS\', \'value\': \'["\\\\\\\\n"]\'}, {\'name\': \'MAX_TOKENS\', \'value\': \'1\'}, {\'name\': \'MAX_BATCH_SIZE\', \'value\': \'128\'}, {\'name\': \'URL_ROUTE\', \'value\': \'GPT-J-6B-lit-v2\'}, {\'name\': \'OBJ_ACCESS_KEY_ID\', \'value\': \'LETMTTRMLFFAMTBK\'}, {\'name\': \'OBJ_SECRET_ACCESS_KEY\', \'value\': \'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\'}, {\'name\': \'OBJ_ENDPOINT\', \'value\': \'https://accel-object.ord1.coreweave.com\'}, {\'name\': \'TENSORIZER_URI\', \'value\': \'s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29\'}, {\'name\': \'RESERVE_MEMORY\', \'value\': \'2048\'}, {\'name\': \'DOWNLOAD_TO_LOCAL\', \'value\': \'/dev/shm/model_cache\'}, {\'name\': \'NUM_GPUS\', \'value\': \'1\'}, {\'name\': \'MK1_MKML_LICENSE_KEY\', \'valueFrom\': {\'secretKeyRef\': {\'key\': \'key\', \'name\': \'mkml-license-key\'}}}], \'image\': \'gcr.io/chai-959f8/chai-guanaco/mkml:mkml_v0.11.12_dg\', \'imagePullPolicy\': \'IfNotPresent\', \'name\': \'kserve-container\', \'readinessProbe\': {\'exec\': {\'command\': [\'cat\', \'/tmp/ready\']}, \'failureThreshold\': 1, \'initialDelaySeconds\': 10, \'periodSeconds\': 10, \'successThreshold\': 1, \'timeoutSeconds\': 5}, \'resources\': {\'limits\': {\'cpu\': \'2\', \'memory\': \'12Gi\', \'nvidia.com/gpu\': \'1\'}, \'requests\': {\'cpu\': \'2\', \'memory\': \'12Gi\', \'nvidia.com/gpu\': \'1\'}}, \'volumeMounts\': [{\'mountPath\': \'/dev/shm\', \'name\': \'shared-memory-cache\'}]}], \'imagePullSecrets\': [{\'name\': \'docker-creds\'}], \'maxReplicas\': 1, \'minReplicas\': 1, \'timeout\': 60, \'volumes\': [{\'emptyDir\': {\'medium\': \'Memory\'}, \'name\': \'shared-memory-cache\'}]}}, \'status\': {\'components\': {\'predictor\': {\'latestCreatedRevision\': \'chaiml-llama-8b-multih-78780-v29-profiler-predictor-00001\'}}, \'conditions\': [{\'lastTransitionTime\': \'2025-02-14T02:28:09Z\', \'reason\': \'PredictorConfigurationReady not ready\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'LatestDeploymentReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:28:09Z\', \'message\': \'Revision "chaiml-llama-8b-multih-78780-v29-profiler-predictor-00001" failed with message: Container failed with: quantization_profile=s0, all_reduce_profile=None, kv_cache_profile=None, calibration_samples=-1, sampling=SamplingParameters(temperature=1.0, top_p=1.0, min_p=0.0, top_k=40, max_input_tokens=256, max_tokens=1, stop=[\\\'\\\\n\\\'], eos_token_ids=[], frequency_penalty=0.0, presence_penalty=0.0, reward_enabled=True, num_samples=1, reward_max_token_input=256, drop_incomplete_sentences=True, profile=False), url_route=GPT-J-6B-lit-v2, tensorizer_uri=s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29, s3_creds=S3Credentials(s3_access_key_id=\\\'LETMTTRMLFFAMTBK\\\', s3_secret_access_key=\\\'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\\\', s3_endpoint=\\\'https://accel-object.ord1.coreweave.com\\\', s3_uncached_endpoint=\\\'https://object.ord1.coreweave.com\\\'), local_folder=/dev/shm/model_cache)\\n[INFO] Initialized device rank 0\\nTraceback (most recent call last):\\n  File "/code/mkml_inference_service/main.py", line 95, in <module>\\n    model.load()\\n  File "/code/mkml_inference_service/main.py", line 31, in load\\n    self.engine = mkml_backend.AsyncInferenceService.from_folder(settings, settings.local_folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 49, in from_folder\\n    return service._from_folder(settings, folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 71, in _from_folder\\n    engine = mkml.ModelForInference.from_pretrained(\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/inference.py", line 66, in from_pretrained\\n    manifold = TensorManifold(model_path, tensor_parallel_size, batching_config, profile, s3_config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 152, in __init__\\n    self.model_actor.load(model_path, profile)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 63, in load\\n    Factory = get_model_factory(self.config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/instrument.py", line 65, in get_model_factory\\n    raise NotImplementedError(config.architectures)\\nNotImplementedError: [\\\'MultiHeadLlamaClassifier\\\']\\n.\', \'reason\': \'RevisionFailed\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'PredictorConfigurationReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:28:09Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'status\': \'False\', \'type\': \'PredictorReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:28:09Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'PredictorRouteReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:28:09Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'status\': \'False\', \'type\': \'Ready\'}, {\'lastTransitionTime\': \'2025-02-14T02:28:09Z\', \'reason\': \'PredictorRouteReady not ready\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'RoutesReady\'}], \'modelStatus\': {\'lastFailureInfo\': {\'exitCode\': 1, \'message\': \'quantization_profile=s0, all_reduce_profile=None, kv_cache_profile=None, calibration_samples=-1, sampling=SamplingParameters(temperature=1.0, top_p=1.0, min_p=0.0, top_k=40, max_input_tokens=256, max_tokens=1, stop=[\\\'\\\\n\\\'], eos_token_ids=[], frequency_penalty=0.0, presence_penalty=0.0, reward_enabled=True, num_samples=1, reward_max_token_input=256, drop_incomplete_sentences=True, profile=False), url_route=GPT-J-6B-lit-v2, tensorizer_uri=s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29, s3_creds=S3Credentials(s3_access_key_id=\\\'LETMTTRMLFFAMTBK\\\', s3_secret_access_key=\\\'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\\\', s3_endpoint=\\\'https://accel-object.ord1.coreweave.com\\\', s3_uncached_endpoint=\\\'https://object.ord1.coreweave.com\\\'), local_folder=/dev/shm/model_cache)\\n[INFO] Initialized device rank 0\\nTraceback (most recent call last):\\n  File "/code/mkml_inference_service/main.py", line 95, in <module>\\n    model.load()\\n  File "/code/mkml_inference_service/main.py", line 31, in load\\n    self.engine = mkml_backend.AsyncInferenceService.from_folder(settings, settings.local_folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 49, in from_folder\\n    return service._from_folder(settings, folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 71, in _from_folder\\n    engine = mkml.ModelForInference.from_pretrained(\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/inference.py", line 66, in from_pretrained\\n    manifold = TensorManifold(model_path, tensor_parallel_size, batching_config, profile, s3_config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 152, in __init__\\n    self.model_actor.load(model_path, profile)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 63, in load\\n    Factory = get_model_factory(self.config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/instrument.py", line 65, in get_model_factory\\n    raise NotImplementedError(config.architectures)\\nNotImplementedError: [\\\'MultiHeadLlamaClassifier\\\']\\n\', \'reason\': \'ModelLoadFailed\'}, \'states\': {\'activeModelState\': \'\', \'targetModelState\': \'FailedToLoad\'}, \'transitionStatus\': \'BlockedByFailedLoad\'}, \'observedGeneration\': 1}}')
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.13s
Shutdown handler de-registered
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
%s, retrying in %s seconds...
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
%s, retrying in %s seconds...
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
clean up pipeline due to error=DeploymentError('Timeout to start the InferenceService chaiml-llama-8b-multih-78780-v29-profiler.                                The InferenceService is as following: {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'kind\': \'InferenceService\', \'metadata\': {\'annotations\': {\'autoscaling.knative.dev/class\': \'hpa.autoscaling.knative.dev\', \'autoscaling.knative.dev/container-concurrency-target-percentage\': \'70\', \'autoscaling.knative.dev/initial-scale\': \'1\', \'autoscaling.knative.dev/max-scale-down-rate\': \'1.1\', \'autoscaling.knative.dev/max-scale-up-rate\': \'2\', \'autoscaling.knative.dev/metric\': \'mean_pod_latency_ms_v2\', \'autoscaling.knative.dev/panic-threshold-percentage\': \'650\', \'autoscaling.knative.dev/panic-window-percentage\': \'35\', \'autoscaling.knative.dev/scale-down-delay\': \'30s\', \'autoscaling.knative.dev/scale-to-zero-grace-period\': \'10m\', \'autoscaling.knative.dev/stable-window\': \'180s\', \'autoscaling.knative.dev/target\': \'3700\', \'autoscaling.knative.dev/target-burst-capacity\': \'-1\', \'autoscaling.knative.dev/tick-interval\': \'15s\', \'features.knative.dev/http-full-duplex\': \'Enabled\', \'networking.knative.dev/ingress-class\': \'istio.ingress.networking.knative.dev\'}, \'creationTimestamp\': \'2025-02-14T02:57:22Z\', \'finalizers\': [\'inferenceservice.finalizers\'], \'generation\': 1, \'labels\': {\'knative.coreweave.cloud/ingress\': \'istio.ingress.networking.knative.dev\', \'prometheus.k.chaiverse.com\': \'true\', \'qos.coreweave.cloud/latency\': \'low\'}, \'managedFields\': [{\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:metadata\': {\'f:annotations\': {\'.\': {}, \'f:autoscaling.knative.dev/class\': {}, \'f:autoscaling.knative.dev/container-concurrency-target-percentage\': {}, \'f:autoscaling.knative.dev/initial-scale\': {}, \'f:autoscaling.knative.dev/max-scale-down-rate\': {}, \'f:autoscaling.knative.dev/max-scale-up-rate\': {}, \'f:autoscaling.knative.dev/metric\': {}, \'f:autoscaling.knative.dev/panic-threshold-percentage\': {}, \'f:autoscaling.knative.dev/panic-window-percentage\': {}, \'f:autoscaling.knative.dev/scale-down-delay\': {}, \'f:autoscaling.knative.dev/scale-to-zero-grace-period\': {}, \'f:autoscaling.knative.dev/stable-window\': {}, \'f:autoscaling.knative.dev/target\': {}, \'f:autoscaling.knative.dev/target-burst-capacity\': {}, \'f:autoscaling.knative.dev/tick-interval\': {}, \'f:features.knative.dev/http-full-duplex\': {}, \'f:networking.knative.dev/ingress-class\': {}}, \'f:labels\': {\'.\': {}, \'f:knative.coreweave.cloud/ingress\': {}, \'f:prometheus.k.chaiverse.com\': {}, \'f:qos.coreweave.cloud/latency\': {}}}, \'f:spec\': {\'.\': {}, \'f:predictor\': {\'.\': {}, \'f:affinity\': {\'.\': {}, \'f:nodeAffinity\': {\'.\': {}, \'f:tion\': {}, \'f:requiredDuringSchedulingIgnoredDuringExecution\': {}}}, \'f:containerConcurrency\': {}, \'f:containers\': {}, \'f:imagePullSecrets\': {}, \'f:maxReplicas\': {}, \'f:minReplicas\': {}, \'f:timeout\': {}, \'f:volumes\': {}}}}, \'manager\': \'OpenAPI-Generator\', \'operation\': \'Update\', \'time\': \'2025-02-14T02:57:22Z\'}, {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:metadata\': {\'f:finalizers\': {\'.\': {}, \'v:"inferenceservice.finalizers"\': {}}}}, \'manager\': \'manager\', \'operation\': \'Update\', \'time\': \'2025-02-14T02:57:22Z\'}, {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:status\': {\'.\': {}, \'f:components\': {\'.\': {}, \'f:predictor\': {\'.\': {}, \'f:latestCreatedRevision\': {}}}, \'f:conditions\': {}, \'f:modelStatus\': {\'.\': {}, \'f:lastFailureInfo\': {\'.\': {}, \'f:exitCode\': {}, \'f:message\': {}, \'f:reason\': {}}, \'f:states\': {\'.\': {}, \'f:activeModelState\': {}, \'f:targetModelState\': {}}, \'f:transitionStatus\': {}}, \'f:observedGeneration\': {}}}, \'manager\': \'manager\', \'operation\': \'Update\', \'subresource\': \'status\', \'time\': \'2025-02-14T02:57:54Z\'}], \'name\': \'chaiml-llama-8b-multih-78780-v29-profiler\', \'namespace\': \'tenant-chaiml-guanaco\', \'resourceVersion\': \'275920492\', \'uid\': \'69b47d60-7153-4c30-8aec-1954520e8faa\'}, \'spec\': {\'predictor\': {\'affinity\': {\'nodeAffinity\': {\'tion\': [{\'preference\': {\'matchExpressions\': [{\'key\': \'topology.kubernetes.io/region\', \'operator\': \'In\', \'values\': [\'ORD1\']}]}, \'weight\': 5}], \'requiredDuringSchedulingIgnoredDuringExecution\': {\'nodeSelectorTerms\': [{\'matchExpressions\': [{\'key\': \'gpu.nvidia.com/class\', \'operator\': \'In\', \'values\': [\'RTX_A5000\']}]}]}}}, \'containerConcurrency\': 0, \'containers\': [{\'env\': [{\'name\': \'MAX_TOKEN_INPUT\', \'value\': \'256\'}, {\'name\': \'BEST_OF\', \'value\': \'1\'}, {\'name\': \'TEMPERATURE\', \'value\': \'1.0\'}, {\'name\': \'PRESENCE_PENALTY\', \'value\': \'0.0\'}, {\'name\': \'FREQUENCY_PENALTY\', \'value\': \'0.0\'}, {\'name\': \'TOP_P\', \'value\': \'1.0\'}, {\'name\': \'MIN_P\', \'value\': \'0.0\'}, {\'name\': \'TOP_K\', \'value\': \'40\'}, {\'name\': \'STOPPING_WORDS\', \'value\': \'["\\\\\\\\n"]\'}, {\'name\': \'MAX_TOKENS\', \'value\': \'1\'}, {\'name\': \'MAX_BATCH_SIZE\', \'value\': \'128\'}, {\'name\': \'URL_ROUTE\', \'value\': \'GPT-J-6B-lit-v2\'}, {\'name\': \'OBJ_ACCESS_KEY_ID\', \'value\': \'LETMTTRMLFFAMTBK\'}, {\'name\': \'OBJ_SECRET_ACCESS_KEY\', \'value\': \'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\'}, {\'name\': \'OBJ_ENDPOINT\', \'value\': \'https://accel-object.ord1.coreweave.com\'}, {\'name\': \'TENSORIZER_URI\', \'value\': \'s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29\'}, {\'name\': \'RESERVE_MEMORY\', \'value\': \'2048\'}, {\'name\': \'DOWNLOAD_TO_LOCAL\', \'value\': \'/dev/shm/model_cache\'}, {\'name\': \'NUM_GPUS\', \'value\': \'1\'}, {\'name\': \'MK1_MKML_LICENSE_KEY\', \'valueFrom\': {\'secretKeyRef\': {\'key\': \'key\', \'name\': \'mkml-license-key\'}}}], \'image\': \'gcr.io/chai-959f8/chai-guanaco/mkml:mkml_v0.11.12_dg\', \'imagePullPolicy\': \'IfNotPresent\', \'name\': \'kserve-container\', \'readinessProbe\': {\'exec\': {\'command\': [\'cat\', \'/tmp/ready\']}, \'failureThreshold\': 1, \'initialDelaySeconds\': 10, \'periodSeconds\': 10, \'successThreshold\': 1, \'timeoutSeconds\': 5}, \'resources\': {\'limits\': {\'cpu\': \'2\', \'memory\': \'12Gi\', \'nvidia.com/gpu\': \'1\'}, \'requests\': {\'cpu\': \'2\', \'memory\': \'12Gi\', \'nvidia.com/gpu\': \'1\'}}, \'volumeMounts\': [{\'mountPath\': \'/dev/shm\', \'name\': \'shared-memory-cache\'}]}], \'imagePullSecrets\': [{\'name\': \'docker-creds\'}], \'maxReplicas\': 1, \'minReplicas\': 1, \'timeout\': 60, \'volumes\': [{\'emptyDir\': {\'medium\': \'Memory\'}, \'name\': \'shared-memory-cache\'}]}}, \'status\': {\'components\': {\'predictor\': {\'latestCreatedRevision\': \'chaiml-llama-8b-multih-78780-v29-profiler-predictor-00001\'}}, \'conditions\': [{\'lastTransitionTime\': \'2025-02-14T02:57:54Z\', \'reason\': \'PredictorConfigurationReady not ready\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'LatestDeploymentReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:57:54Z\', \'message\': \'Revision "chaiml-llama-8b-multih-78780-v29-profiler-predictor-00001" failed with message: Container failed with: quantization_profile=s0, all_reduce_profile=None, kv_cache_profile=None, calibration_samples=-1, sampling=SamplingParameters(temperature=1.0, top_p=1.0, min_p=0.0, top_k=40, max_input_tokens=256, max_tokens=1, stop=[\\\'\\\\n\\\'], eos_token_ids=[], frequency_penalty=0.0, presence_penalty=0.0, reward_enabled=True, num_samples=1, reward_max_token_input=256, drop_incomplete_sentences=True, profile=False), url_route=GPT-J-6B-lit-v2, tensorizer_uri=s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29, s3_creds=S3Credentials(s3_access_key_id=\\\'LETMTTRMLFFAMTBK\\\', s3_secret_access_key=\\\'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\\\', s3_endpoint=\\\'https://accel-object.ord1.coreweave.com\\\', s3_uncached_endpoint=\\\'https://object.ord1.coreweave.com\\\'), local_folder=/dev/shm/model_cache)\\n[INFO] Initialized device rank 0\\nTraceback (most recent call last):\\n  File "/code/mkml_inference_service/main.py", line 95, in <module>\\n    model.load()\\n  File "/code/mkml_inference_service/main.py", line 31, in load\\n    self.engine = mkml_backend.AsyncInferenceService.from_folder(settings, settings.local_folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 49, in from_folder\\n    return service._from_folder(settings, folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 71, in _from_folder\\n    engine = mkml.ModelForInference.from_pretrained(\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/inference.py", line 66, in from_pretrained\\n    manifold = TensorManifold(model_path, tensor_parallel_size, batching_config, profile, s3_config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 152, in __init__\\n    self.model_actor.load(model_path, profile)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 63, in load\\n    Factory = get_model_factory(self.config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/instrument.py", line 65, in get_model_factory\\n    raise NotImplementedError(config.architectures)\\nNotImplementedError: [\\\'MultiHeadLlamaClassifier\\\']\\n.\', \'reason\': \'RevisionFailed\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'PredictorConfigurationReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:57:54Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'status\': \'False\', \'type\': \'PredictorReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:57:54Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'PredictorRouteReady\'}, {\'lastTransitionTime\': \'2025-02-14T02:57:54Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'status\': \'False\', \'type\': \'Ready\'}, {\'lastTransitionTime\': \'2025-02-14T02:57:54Z\', \'reason\': \'PredictorRouteReady not ready\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'RoutesReady\'}], \'modelStatus\': {\'lastFailureInfo\': {\'exitCode\': 1, \'message\': \'quantization_profile=s0, all_reduce_profile=None, kv_cache_profile=None, calibration_samples=-1, sampling=SamplingParameters(temperature=1.0, top_p=1.0, min_p=0.0, top_k=40, max_input_tokens=256, max_tokens=1, stop=[\\\'\\\\n\\\'], eos_token_ids=[], frequency_penalty=0.0, presence_penalty=0.0, reward_enabled=True, num_samples=1, reward_max_token_input=256, drop_incomplete_sentences=True, profile=False), url_route=GPT-J-6B-lit-v2, tensorizer_uri=s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29, s3_creds=S3Credentials(s3_access_key_id=\\\'LETMTTRMLFFAMTBK\\\', s3_secret_access_key=\\\'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\\\', s3_endpoint=\\\'https://accel-object.ord1.coreweave.com\\\', s3_uncached_endpoint=\\\'https://object.ord1.coreweave.com\\\'), local_folder=/dev/shm/model_cache)\\n[INFO] Initialized device rank 0\\nTraceback (most recent call last):\\n  File "/code/mkml_inference_service/main.py", line 95, in <module>\\n    model.load()\\n  File "/code/mkml_inference_service/main.py", line 31, in load\\n    self.engine = mkml_backend.AsyncInferenceService.from_folder(settings, settings.local_folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 49, in from_folder\\n    return service._from_folder(settings, folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 71, in _from_folder\\n    engine = mkml.ModelForInference.from_pretrained(\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/inference.py", line 66, in from_pretrained\\n    manifold = TensorManifold(model_path, tensor_parallel_size, batching_config, profile, s3_config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 152, in __init__\\n    self.model_actor.load(model_path, profile)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 63, in load\\n    Factory = get_model_factory(self.config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/instrument.py", line 65, in get_model_factory\\n    raise NotImplementedError(config.architectures)\\nNotImplementedError: [\\\'MultiHeadLlamaClassifier\\\']\\n\', \'reason\': \'ModelLoadFailed\'}, \'states\': {\'activeModelState\': \'\', \'targetModelState\': \'FailedToLoad\'}, \'transitionStatus\': \'BlockedByFailedLoad\'}, \'observedGeneration\': 1}}')
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.14s
Shutdown handler de-registered
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.17s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
%s, retrying in %s seconds...
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
%s, retrying in %s seconds...
Creating inference service chaiml-llama-8b-multih-78780-v29-profiler
Waiting for inference service chaiml-llama-8b-multih-78780-v29-profiler to be ready
Tearing down inference service chaiml-llama-8b-multih-78780-v29-profiler
clean up pipeline due to error=DeploymentError('Timeout to start the InferenceService chaiml-llama-8b-multih-78780-v29-profiler.                                The InferenceService is as following: {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'kind\': \'InferenceService\', \'metadata\': {\'annotations\': {\'autoscaling.knative.dev/class\': \'hpa.autoscaling.knative.dev\', \'autoscaling.knative.dev/container-concurrency-target-percentage\': \'70\', \'autoscaling.knative.dev/initial-scale\': \'1\', \'autoscaling.knative.dev/max-scale-down-rate\': \'1.1\', \'autoscaling.knative.dev/max-scale-up-rate\': \'2\', \'autoscaling.knative.dev/metric\': \'mean_pod_latency_ms_v2\', \'autoscaling.knative.dev/panic-threshold-percentage\': \'650\', \'autoscaling.knative.dev/panic-window-percentage\': \'35\', \'autoscaling.knative.dev/scale-down-delay\': \'30s\', \'autoscaling.knative.dev/scale-to-zero-grace-period\': \'10m\', \'autoscaling.knative.dev/stable-window\': \'180s\', \'autoscaling.knative.dev/target\': \'3700\', \'autoscaling.knative.dev/target-burst-capacity\': \'-1\', \'autoscaling.knative.dev/tick-interval\': \'15s\', \'features.knative.dev/http-full-duplex\': \'Enabled\', \'networking.knative.dev/ingress-class\': \'istio.ingress.networking.knative.dev\'}, \'creationTimestamp\': \'2025-02-14T03:28:15Z\', \'finalizers\': [\'inferenceservice.finalizers\'], \'generation\': 1, \'labels\': {\'knative.coreweave.cloud/ingress\': \'istio.ingress.networking.knative.dev\', \'prometheus.k.chaiverse.com\': \'true\', \'qos.coreweave.cloud/latency\': \'low\'}, \'managedFields\': [{\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:metadata\': {\'f:annotations\': {\'.\': {}, \'f:autoscaling.knative.dev/class\': {}, \'f:autoscaling.knative.dev/container-concurrency-target-percentage\': {}, \'f:autoscaling.knative.dev/initial-scale\': {}, \'f:autoscaling.knative.dev/max-scale-down-rate\': {}, \'f:autoscaling.knative.dev/max-scale-up-rate\': {}, \'f:autoscaling.knative.dev/metric\': {}, \'f:autoscaling.knative.dev/panic-threshold-percentage\': {}, \'f:autoscaling.knative.dev/panic-window-percentage\': {}, \'f:autoscaling.knative.dev/scale-down-delay\': {}, \'f:autoscaling.knative.dev/scale-to-zero-grace-period\': {}, \'f:autoscaling.knative.dev/stable-window\': {}, \'f:autoscaling.knative.dev/target\': {}, \'f:autoscaling.knative.dev/target-burst-capacity\': {}, \'f:autoscaling.knative.dev/tick-interval\': {}, \'f:features.knative.dev/http-full-duplex\': {}, \'f:networking.knative.dev/ingress-class\': {}}, \'f:labels\': {\'.\': {}, \'f:knative.coreweave.cloud/ingress\': {}, \'f:prometheus.k.chaiverse.com\': {}, \'f:qos.coreweave.cloud/latency\': {}}}, \'f:spec\': {\'.\': {}, \'f:predictor\': {\'.\': {}, \'f:affinity\': {\'.\': {}, \'f:nodeAffinity\': {\'.\': {}, \'f:tion\': {}, \'f:requiredDuringSchedulingIgnoredDuringExecution\': {}}}, \'f:containerConcurrency\': {}, \'f:containers\': {}, \'f:imagePullSecrets\': {}, \'f:maxReplicas\': {}, \'f:minReplicas\': {}, \'f:timeout\': {}, \'f:volumes\': {}}}}, \'manager\': \'OpenAPI-Generator\', \'operation\': \'Update\', \'time\': \'2025-02-14T03:28:15Z\'}, {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:metadata\': {\'f:finalizers\': {\'.\': {}, \'v:"inferenceservice.finalizers"\': {}}}}, \'manager\': \'manager\', \'operation\': \'Update\', \'time\': \'2025-02-14T03:28:15Z\'}, {\'apiVersion\': \'serving.kserve.io/v1beta1\', \'fieldsType\': \'FieldsV1\', \'fieldsV1\': {\'f:status\': {\'.\': {}, \'f:components\': {\'.\': {}, \'f:predictor\': {\'.\': {}, \'f:latestCreatedRevision\': {}}}, \'f:conditions\': {}, \'f:modelStatus\': {\'.\': {}, \'f:lastFailureInfo\': {\'.\': {}, \'f:exitCode\': {}, \'f:message\': {}, \'f:reason\': {}}, \'f:states\': {\'.\': {}, \'f:activeModelState\': {}, \'f:targetModelState\': {}}, \'f:transitionStatus\': {}}, \'f:observedGeneration\': {}}}, \'manager\': \'manager\', \'operation\': \'Update\', \'subresource\': \'status\', \'time\': \'2025-02-14T03:30:36Z\'}], \'name\': \'chaiml-llama-8b-multih-78780-v29-profiler\', \'namespace\': \'tenant-chaiml-guanaco\', \'resourceVersion\': \'275950817\', \'uid\': \'98d10b5a-f9dd-4cb6-8836-649d969be0e5\'}, \'spec\': {\'predictor\': {\'affinity\': {\'nodeAffinity\': {\'tion\': [{\'preference\': {\'matchExpressions\': [{\'key\': \'topology.kubernetes.io/region\', \'operator\': \'In\', \'values\': [\'ORD1\']}]}, \'weight\': 5}], \'requiredDuringSchedulingIgnoredDuringExecution\': {\'nodeSelectorTerms\': [{\'matchExpressions\': [{\'key\': \'gpu.nvidia.com/class\', \'operator\': \'In\', \'values\': [\'RTX_A5000\']}]}]}}}, \'containerConcurrency\': 0, \'containers\': [{\'env\': [{\'name\': \'MAX_TOKEN_INPUT\', \'value\': \'256\'}, {\'name\': \'BEST_OF\', \'value\': \'1\'}, {\'name\': \'TEMPERATURE\', \'value\': \'1.0\'}, {\'name\': \'PRESENCE_PENALTY\', \'value\': \'0.0\'}, {\'name\': \'FREQUENCY_PENALTY\', \'value\': \'0.0\'}, {\'name\': \'TOP_P\', \'value\': \'1.0\'}, {\'name\': \'MIN_P\', \'value\': \'0.0\'}, {\'name\': \'TOP_K\', \'value\': \'40\'}, {\'name\': \'STOPPING_WORDS\', \'value\': \'["\\\\\\\\n"]\'}, {\'name\': \'MAX_TOKENS\', \'value\': \'1\'}, {\'name\': \'MAX_BATCH_SIZE\', \'value\': \'128\'}, {\'name\': \'URL_ROUTE\', \'value\': \'GPT-J-6B-lit-v2\'}, {\'name\': \'OBJ_ACCESS_KEY_ID\', \'value\': \'LETMTTRMLFFAMTBK\'}, {\'name\': \'OBJ_SECRET_ACCESS_KEY\', \'value\': \'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\'}, {\'name\': \'OBJ_ENDPOINT\', \'value\': \'https://accel-object.ord1.coreweave.com\'}, {\'name\': \'TENSORIZER_URI\', \'value\': \'s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29\'}, {\'name\': \'RESERVE_MEMORY\', \'value\': \'2048\'}, {\'name\': \'DOWNLOAD_TO_LOCAL\', \'value\': \'/dev/shm/model_cache\'}, {\'name\': \'NUM_GPUS\', \'value\': \'1\'}, {\'name\': \'MK1_MKML_LICENSE_KEY\', \'valueFrom\': {\'secretKeyRef\': {\'key\': \'key\', \'name\': \'mkml-license-key\'}}}], \'image\': \'gcr.io/chai-959f8/chai-guanaco/mkml:mkml_v0.11.12_dg\', \'imagePullPolicy\': \'IfNotPresent\', \'name\': \'kserve-container\', \'readinessProbe\': {\'exec\': {\'command\': [\'cat\', \'/tmp/ready\']}, \'failureThreshold\': 1, \'initialDelaySeconds\': 10, \'periodSeconds\': 10, \'successThreshold\': 1, \'timeoutSeconds\': 5}, \'resources\': {\'limits\': {\'cpu\': \'2\', \'memory\': \'12Gi\', \'nvidia.com/gpu\': \'1\'}, \'requests\': {\'cpu\': \'2\', \'memory\': \'12Gi\', \'nvidia.com/gpu\': \'1\'}}, \'volumeMounts\': [{\'mountPath\': \'/dev/shm\', \'name\': \'shared-memory-cache\'}]}], \'imagePullSecrets\': [{\'name\': \'docker-creds\'}], \'maxReplicas\': 1, \'minReplicas\': 1, \'timeout\': 60, \'volumes\': [{\'emptyDir\': {\'medium\': \'Memory\'}, \'name\': \'shared-memory-cache\'}]}}, \'status\': {\'components\': {\'predictor\': {\'latestCreatedRevision\': \'chaiml-llama-8b-multih-78780-v29-profiler-predictor-00001\'}}, \'conditions\': [{\'lastTransitionTime\': \'2025-02-14T03:30:36Z\', \'reason\': \'PredictorConfigurationReady not ready\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'LatestDeploymentReady\'}, {\'lastTransitionTime\': \'2025-02-14T03:30:36Z\', \'message\': \'Revision "chaiml-llama-8b-multih-78780-v29-profiler-predictor-00001" failed with message: Container failed with: quantization_profile=s0, all_reduce_profile=None, kv_cache_profile=None, calibration_samples=-1, sampling=SamplingParameters(temperature=1.0, top_p=1.0, min_p=0.0, top_k=40, max_input_tokens=256, max_tokens=1, stop=[\\\'\\\\n\\\'], eos_token_ids=[], frequency_penalty=0.0, presence_penalty=0.0, reward_enabled=True, num_samples=1, reward_max_token_input=256, drop_incomplete_sentences=True, profile=False), url_route=GPT-J-6B-lit-v2, tensorizer_uri=s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29, s3_creds=S3Credentials(s3_access_key_id=\\\'LETMTTRMLFFAMTBK\\\', s3_secret_access_key=\\\'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\\\', s3_endpoint=\\\'https://accel-object.ord1.coreweave.com\\\', s3_uncached_endpoint=\\\'https://object.ord1.coreweave.com\\\'), local_folder=/dev/shm/model_cache)\\n[INFO] Initialized device rank 0\\nTraceback (most recent call last):\\n  File "/code/mkml_inference_service/main.py", line 95, in <module>\\n    model.load()\\n  File "/code/mkml_inference_service/main.py", line 31, in load\\n    self.engine = mkml_backend.AsyncInferenceService.from_folder(settings, settings.local_folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 49, in from_folder\\n    return service._from_folder(settings, folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 71, in _from_folder\\n    engine = mkml.ModelForInference.from_pretrained(\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/inference.py", line 66, in from_pretrained\\n    manifold = TensorManifold(model_path, tensor_parallel_size, batching_config, profile, s3_config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 152, in __init__\\n    self.model_actor.load(model_path, profile)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 63, in load\\n    Factory = get_model_factory(self.config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/instrument.py", line 65, in get_model_factory\\n    raise NotImplementedError(config.architectures)\\nNotImplementedError: [\\\'MultiHeadLlamaClassifier\\\']\\n.\', \'reason\': \'RevisionFailed\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'PredictorConfigurationReady\'}, {\'lastTransitionTime\': \'2025-02-14T03:30:36Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'status\': \'False\', \'type\': \'PredictorReady\'}, {\'lastTransitionTime\': \'2025-02-14T03:30:36Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'PredictorRouteReady\'}, {\'lastTransitionTime\': \'2025-02-14T03:30:36Z\', \'message\': \'Configuration "chaiml-llama-8b-multih-78780-v29-profiler-predictor" does not have any ready Revision.\', \'reason\': \'RevisionMissing\', \'status\': \'False\', \'type\': \'Ready\'}, {\'lastTransitionTime\': \'2025-02-14T03:30:36Z\', \'reason\': \'PredictorRouteReady not ready\', \'severity\': \'Info\', \'status\': \'False\', \'type\': \'RoutesReady\'}], \'modelStatus\': {\'lastFailureInfo\': {\'exitCode\': 1, \'message\': \'quantization_profile=s0, all_reduce_profile=None, kv_cache_profile=None, calibration_samples=-1, sampling=SamplingParameters(temperature=1.0, top_p=1.0, min_p=0.0, top_k=40, max_input_tokens=256, max_tokens=1, stop=[\\\'\\\\n\\\'], eos_token_ids=[], frequency_penalty=0.0, presence_penalty=0.0, reward_enabled=True, num_samples=1, reward_max_token_input=256, drop_incomplete_sentences=True, profile=False), url_route=GPT-J-6B-lit-v2, tensorizer_uri=s3://guanaco-mkml-models/chaiml-llama-8b-multih-78780-v29, s3_creds=S3Credentials(s3_access_key_id=\\\'LETMTTRMLFFAMTBK\\\', s3_secret_access_key=\\\'VwwZaqefOOoaouNxUk03oUmK9pVEfruJhjBHPGdgycK\\\', s3_endpoint=\\\'https://accel-object.ord1.coreweave.com\\\', s3_uncached_endpoint=\\\'https://object.ord1.coreweave.com\\\'), local_folder=/dev/shm/model_cache)\\n[INFO] Initialized device rank 0\\nTraceback (most recent call last):\\n  File "/code/mkml_inference_service/main.py", line 95, in <module>\\n    model.load()\\n  File "/code/mkml_inference_service/main.py", line 31, in load\\n    self.engine = mkml_backend.AsyncInferenceService.from_folder(settings, settings.local_folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 49, in from_folder\\n    return service._from_folder(settings, folder)\\n  File "/code/mkml_inference_service/mkml_backend.py", line 71, in _from_folder\\n    engine = mkml.ModelForInference.from_pretrained(\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/inference.py", line 66, in from_pretrained\\n    manifold = TensorManifold(model_path, tensor_parallel_size, batching_config, profile, s3_config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 152, in __init__\\n    self.model_actor.load(model_path, profile)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/manifold.py", line 63, in load\\n    Factory = get_model_factory(self.config)\\n  File "/opt/conda/lib/python3.10/site-packages/mk1/flywheel/instrument.py", line 65, in get_model_factory\\n    raise NotImplementedError(config.architectures)\\nNotImplementedError: [\\\'MultiHeadLlamaClassifier\\\']\\n\', \'reason\': \'ModelLoadFailed\'}, \'states\': {\'activeModelState\': \'\', \'targetModelState\': \'FailedToLoad\'}, \'transitionStatus\': \'BlockedByFailedLoad\'}, \'observedGeneration\': 1}}')
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.17s
Shutdown handler de-registered
chaiml-llama-8b-multih_78780_v29 status is now inactive due to auto deactivation removed underperforming models
chaiml-llama-8b-multih_78780_v29 status is now torndown due to DeploymentManager action
chaiml-llama-8b-multih_78780_v29 status is now torndown due to DeploymentManager action
chaiml-llama-8b-multih_78780_v29 status is now torndown due to DeploymentManager action
ChatRequest
Generation Params
Prompt Formatter
Chat History
ChatMessage 1