nsthorat-lilac commited on
Commit
86fe272
1 Parent(s): 02f502d

Push to HF space

Browse files
Files changed (3) hide show
  1. Dockerfile +6 -2
  2. data/lilac.yml +448 -56
  3. docker_start.sh +1 -1
Dockerfile CHANGED
@@ -1,9 +1,12 @@
1
- # NOTE: When we upgrade to 3.11 we can use a slimmer docker image which comes with gcc.
2
- FROM python:3.9-bullseye
3
 
4
  # Allow statements and log messages to immediately appear in the Knative logs
5
  ENV PYTHONUNBUFFERED True
6
 
 
 
 
 
7
  # See: https://huggingface.co/docs/hub/spaces-sdks-docker#permissions
8
  RUN useradd -m -u 1000 user
9
  USER user
@@ -39,4 +42,5 @@ COPY --chown=user docker_start.sh ./
39
  RUN mkdir -p ./data
40
  RUN chown -R user ./data
41
 
 
42
  CMD ["bash", "docker_start.sh"]
 
1
+ FROM python:3.11-slim-bullseye
 
2
 
3
  # Allow statements and log messages to immediately appear in the Knative logs
4
  ENV PYTHONUNBUFFERED True
5
 
6
+ # Adds GCC and other build tools so we can compile hnswlib and other native/C++ deps.
7
+ RUN apt-get update --fix-missing && apt-get install -y --fix-missing build-essential && \
8
+ rm -rf /var/lib/apt/lists/*
9
+
10
  # See: https://huggingface.co/docs/hub/spaces-sdks-docker#permissions
11
  RUN useradd -m -u 1000 user
12
  USER user
 
42
  RUN mkdir -p ./data
43
  RUN chown -R user ./data
44
 
45
+ EXPOSE 8000
46
  CMD ["bash", "docker_start.sh"]
data/lilac.yml CHANGED
@@ -230,10 +230,13 @@ datasets:
230
  signal_name: concept_score
231
  - path: text
232
  signal:
233
- embedding: gte-small
234
- eps: 0.05
235
- min_samples: 5
236
  signal_name: cluster_dbscan
 
 
 
 
 
 
237
  settings:
238
  ui:
239
  media_paths:
@@ -407,10 +410,13 @@ datasets:
407
  signal_name: concept_score
408
  - path: text
409
  signal:
410
- embedding: gte-small
411
- eps: 0.05
412
- min_samples: 5
413
  signal_name: cluster_dbscan
 
 
 
 
 
 
414
  settings:
415
  ui:
416
  media_paths:
@@ -576,10 +582,13 @@ datasets:
576
  signal_name: concept_score
577
  - path: text
578
  signal:
579
- embedding: gte-small
580
- eps: 0.05
581
- min_samples: 5
582
  signal_name: cluster_dbscan
 
 
 
 
 
 
583
  settings:
584
  ui:
585
  media_paths:
@@ -605,6 +614,20 @@ datasets:
605
  - value
606
  - '*'
607
  embedding: gte-small
 
 
 
 
 
 
 
 
 
 
 
 
 
 
608
  signals:
609
  - path: original-instruction
610
  signal:
@@ -2856,9 +2879,6 @@ datasets:
2856
  signal_name: concept_score
2857
  - path: original-instruction
2858
  signal:
2859
- embedding: gte-small
2860
- eps: 0.05
2861
- min_samples: 5
2862
  signal_name: cluster_dbscan
2863
  - path: original-instruction
2864
  signal:
@@ -2926,15 +2946,9 @@ datasets:
2926
  signal_name: concept_score
2927
  - path: original-context
2928
  signal:
2929
- embedding: gte-small
2930
- eps: 0.05
2931
- min_samples: 5
2932
  signal_name: cluster_dbscan
2933
  - path: original-response
2934
  signal:
2935
- embedding: gte-small
2936
- eps: 0.05
2937
- min_samples: 5
2938
  signal_name: cluster_dbscan
2939
  - path: original-response
2940
  signal:
@@ -3000,15 +3014,355 @@ datasets:
3000
  version: null
3001
  draft: main
3002
  signal_name: concept_score
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3003
  - path:
3004
  - new-instruction
3005
  - value
3006
  - '*'
3007
  signal:
3008
  embedding: gte-small
3009
- eps: 0.05
3010
- min_samples: 5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3011
  signal_name: cluster_dbscan
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3012
  - path:
3013
  - new-instruction
3014
  - value
@@ -3103,18 +3457,18 @@ datasets:
3103
  - '*'
3104
  signal:
3105
  embedding: gte-small
3106
- eps: 0.05
3107
- min_samples: 5
3108
- signal_name: cluster_dbscan
3109
  - path:
3110
  - new-response
3111
  - value
3112
  - '*'
3113
  signal:
3114
  embedding: gte-small
3115
- eps: 0.05
3116
- min_samples: 5
3117
- signal_name: cluster_dbscan
3118
  - path:
3119
  - new-response
3120
  - value
@@ -3784,16 +4138,22 @@ datasets:
3784
  signal_name: concept_score
3785
  - path: question
3786
  signal:
3787
- embedding: gte-small
3788
- eps: 0.05
3789
- min_samples: 5
3790
  signal_name: cluster_dbscan
3791
  - path: response
3792
  signal:
3793
- embedding: gte-small
3794
- eps: 0.05
3795
- min_samples: 5
3796
  signal_name: cluster_dbscan
 
 
 
 
 
 
 
 
 
 
 
 
3797
  settings:
3798
  ui:
3799
  media_paths:
@@ -4190,18 +4550,28 @@ datasets:
4190
  - translation
4191
  - es
4192
  signal:
4193
- embedding: gte-small
4194
- eps: 0.05
4195
- min_samples: 5
4196
  signal_name: cluster_dbscan
4197
  - path:
4198
  - translation
4199
  - en
4200
  signal:
4201
- embedding: gte-small
4202
- eps: 0.05
4203
- min_samples: 5
4204
  signal_name: cluster_dbscan
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4205
  settings:
4206
  ui:
4207
  media_paths:
@@ -4370,10 +4740,13 @@ datasets:
4370
  signal_name: concept_score
4371
  - path: lecture
4372
  signal:
4373
- embedding: gte-small
4374
- eps: 0.05
4375
- min_samples: 5
4376
  signal_name: cluster_dbscan
 
 
 
 
 
 
4377
  settings:
4378
  ui:
4379
  media_paths:
@@ -4539,10 +4912,13 @@ datasets:
4539
  signal_name: concept_score
4540
  - path: text
4541
  signal:
4542
- embedding: gte-small
4543
- eps: 0.05
4544
- min_samples: 5
4545
  signal_name: cluster_dbscan
 
 
 
 
 
 
4546
  settings:
4547
  ui:
4548
  media_paths:
@@ -4707,10 +5083,13 @@ datasets:
4707
  signal_name: concept_score
4708
  - path: overview
4709
  signal:
4710
- embedding: gte-small
4711
- eps: 0.05
4712
- min_samples: 5
4713
  signal_name: cluster_dbscan
 
 
 
 
 
 
4714
  settings:
4715
  ui:
4716
  media_paths:
@@ -5026,24 +5405,37 @@ datasets:
5026
  - outline
5027
  - '*'
5028
  signal:
5029
- embedding: gte-small
5030
- eps: 0.05
5031
- min_samples: 5
5032
  signal_name: cluster_dbscan
5033
  - path:
5034
  - concepts
5035
  - '*'
5036
  signal:
5037
- embedding: gte-small
5038
- eps: 0.05
5039
- min_samples: 5
5040
  signal_name: cluster_dbscan
5041
  - path: markdown
5042
  signal:
5043
- embedding: gte-small
5044
- eps: 0.05
5045
- min_samples: 5
5046
  signal_name: cluster_dbscan
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5047
  settings:
5048
  ui:
5049
  media_paths:
 
230
  signal_name: concept_score
231
  - path: text
232
  signal:
 
 
 
233
  signal_name: cluster_dbscan
234
+ - path: text
235
+ signal:
236
+ embedding: gte-small
237
+ min_cluster_size: 5
238
+ umap_n_components: 10
239
+ signal_name: cluster_hdbscan
240
  settings:
241
  ui:
242
  media_paths:
 
410
  signal_name: concept_score
411
  - path: text
412
  signal:
 
 
 
413
  signal_name: cluster_dbscan
414
+ - path: text
415
+ signal:
416
+ embedding: gte-small
417
+ min_cluster_size: 5
418
+ umap_n_components: 10
419
+ signal_name: cluster_hdbscan
420
  settings:
421
  ui:
422
  media_paths:
 
582
  signal_name: concept_score
583
  - path: text
584
  signal:
 
 
 
585
  signal_name: cluster_dbscan
586
+ - path: text
587
+ signal:
588
+ embedding: gte-small
589
+ min_cluster_size: 5
590
+ umap_n_components: 10
591
+ signal_name: cluster_hdbscan
592
  settings:
593
  ui:
594
  media_paths:
 
614
  - value
615
  - '*'
616
  embedding: gte-small
617
+ - path: original-instruction
618
+ embedding: gte-small
619
+ - path: original-response
620
+ embedding: gte-small
621
+ - path:
622
+ - new-instruction
623
+ - value
624
+ - '*'
625
+ embedding: gte-small
626
+ - path:
627
+ - new-response
628
+ - value
629
+ - '*'
630
+ embedding: gte-small
631
  signals:
632
  - path: original-instruction
633
  signal:
 
2879
  signal_name: concept_score
2880
  - path: original-instruction
2881
  signal:
 
 
 
2882
  signal_name: cluster_dbscan
2883
  - path: original-instruction
2884
  signal:
 
2946
  signal_name: concept_score
2947
  - path: original-context
2948
  signal:
 
 
 
2949
  signal_name: cluster_dbscan
2950
  - path: original-response
2951
  signal:
 
 
 
2952
  signal_name: cluster_dbscan
2953
  - path: original-response
2954
  signal:
 
3014
  version: null
3015
  draft: main
3016
  signal_name: concept_score
3017
+ - path:
3018
+ - new-instruction
3019
+ - value
3020
+ - '*'
3021
+ signal:
3022
+ signal_name: cluster_dbscan
3023
+ - path:
3024
+ - new-instruction
3025
+ - value
3026
+ - '*'
3027
+ signal:
3028
+ embedding: gte-small
3029
+ namespace: lilac
3030
+ concept_name: legal-termination
3031
+ version: null
3032
+ draft: main
3033
+ signal_name: concept_score
3034
  - path:
3035
  - new-instruction
3036
  - value
3037
  - '*'
3038
  signal:
3039
  embedding: gte-small
3040
+ namespace: lilac
3041
+ concept_name: negative-sentiment
3042
+ version: null
3043
+ draft: main
3044
+ signal_name: concept_score
3045
+ - path:
3046
+ - new-instruction
3047
+ - value
3048
+ - '*'
3049
+ signal:
3050
+ embedding: gte-small
3051
+ namespace: lilac
3052
+ concept_name: non-english
3053
+ version: null
3054
+ draft: main
3055
+ signal_name: concept_score
3056
+ - path:
3057
+ - new-instruction
3058
+ - value
3059
+ - '*'
3060
+ signal:
3061
+ embedding: gte-small
3062
+ namespace: lilac
3063
+ concept_name: positive-sentiment
3064
+ version: null
3065
+ draft: main
3066
+ signal_name: concept_score
3067
+ - path:
3068
+ - new-instruction
3069
+ - value
3070
+ - '*'
3071
+ signal:
3072
+ embedding: gte-small
3073
+ namespace: lilac
3074
+ concept_name: profanity
3075
+ version: null
3076
+ draft: main
3077
+ signal_name: concept_score
3078
+ - path:
3079
+ - new-instruction
3080
+ - value
3081
+ - '*'
3082
+ signal:
3083
+ embedding: gte-small
3084
+ namespace: lilac
3085
+ concept_name: question
3086
+ version: null
3087
+ draft: main
3088
+ signal_name: concept_score
3089
+ - path:
3090
+ - new-instruction
3091
+ - value
3092
+ - '*'
3093
+ signal:
3094
+ embedding: gte-small
3095
+ namespace: lilac
3096
+ concept_name: source-code
3097
+ version: null
3098
+ draft: main
3099
+ signal_name: concept_score
3100
+ - path:
3101
+ - new-instruction
3102
+ - value
3103
+ - '*'
3104
+ signal:
3105
+ embedding: gte-small
3106
+ namespace: lilac
3107
+ concept_name: toxicity
3108
+ version: null
3109
+ draft: main
3110
+ signal_name: concept_score
3111
+ - path:
3112
+ - new-context
3113
+ - value
3114
+ - '*'
3115
+ signal:
3116
  signal_name: cluster_dbscan
3117
+ - path:
3118
+ - new-response
3119
+ - value
3120
+ - '*'
3121
+ signal:
3122
+ signal_name: cluster_dbscan
3123
+ - path:
3124
+ - new-response
3125
+ - value
3126
+ - '*'
3127
+ signal:
3128
+ embedding: gte-small
3129
+ namespace: lilac
3130
+ concept_name: legal-termination
3131
+ version: null
3132
+ draft: main
3133
+ signal_name: concept_score
3134
+ - path:
3135
+ - new-response
3136
+ - value
3137
+ - '*'
3138
+ signal:
3139
+ embedding: gte-small
3140
+ namespace: lilac
3141
+ concept_name: negative-sentiment
3142
+ version: null
3143
+ draft: main
3144
+ signal_name: concept_score
3145
+ - path:
3146
+ - new-response
3147
+ - value
3148
+ - '*'
3149
+ signal:
3150
+ embedding: gte-small
3151
+ namespace: lilac
3152
+ concept_name: non-english
3153
+ version: null
3154
+ draft: main
3155
+ signal_name: concept_score
3156
+ - path:
3157
+ - new-response
3158
+ - value
3159
+ - '*'
3160
+ signal:
3161
+ embedding: gte-small
3162
+ namespace: lilac
3163
+ concept_name: positive-sentiment
3164
+ version: null
3165
+ draft: main
3166
+ signal_name: concept_score
3167
+ - path:
3168
+ - new-response
3169
+ - value
3170
+ - '*'
3171
+ signal:
3172
+ embedding: gte-small
3173
+ namespace: lilac
3174
+ concept_name: profanity
3175
+ version: null
3176
+ draft: main
3177
+ signal_name: concept_score
3178
+ - path:
3179
+ - new-response
3180
+ - value
3181
+ - '*'
3182
+ signal:
3183
+ embedding: gte-small
3184
+ namespace: lilac
3185
+ concept_name: question
3186
+ version: null
3187
+ draft: main
3188
+ signal_name: concept_score
3189
+ - path:
3190
+ - new-response
3191
+ - value
3192
+ - '*'
3193
+ signal:
3194
+ embedding: gte-small
3195
+ namespace: lilac
3196
+ concept_name: source-code
3197
+ version: null
3198
+ draft: main
3199
+ signal_name: concept_score
3200
+ - path:
3201
+ - new-response
3202
+ - value
3203
+ - '*'
3204
+ signal:
3205
+ embedding: gte-small
3206
+ namespace: lilac
3207
+ concept_name: toxicity
3208
+ version: null
3209
+ draft: main
3210
+ signal_name: concept_score
3211
+ - path: original-instruction
3212
+ signal:
3213
+ embedding: gte-small
3214
+ min_cluster_size: 5
3215
+ umap_n_components: 10
3216
+ signal_name: cluster_hdbscan
3217
+ - path: original-instruction
3218
+ signal:
3219
+ embedding: gte-small
3220
+ namespace: lilac
3221
+ concept_name: legal-termination
3222
+ version: null
3223
+ draft: main
3224
+ signal_name: concept_score
3225
+ - path: original-instruction
3226
+ signal:
3227
+ embedding: gte-small
3228
+ namespace: lilac
3229
+ concept_name: negative-sentiment
3230
+ version: null
3231
+ draft: main
3232
+ signal_name: concept_score
3233
+ - path: original-instruction
3234
+ signal:
3235
+ embedding: gte-small
3236
+ namespace: lilac
3237
+ concept_name: non-english
3238
+ version: null
3239
+ draft: main
3240
+ signal_name: concept_score
3241
+ - path: original-instruction
3242
+ signal:
3243
+ embedding: gte-small
3244
+ namespace: lilac
3245
+ concept_name: positive-sentiment
3246
+ version: null
3247
+ draft: main
3248
+ signal_name: concept_score
3249
+ - path: original-instruction
3250
+ signal:
3251
+ embedding: gte-small
3252
+ namespace: lilac
3253
+ concept_name: profanity
3254
+ version: null
3255
+ draft: main
3256
+ signal_name: concept_score
3257
+ - path: original-instruction
3258
+ signal:
3259
+ embedding: gte-small
3260
+ namespace: lilac
3261
+ concept_name: question
3262
+ version: null
3263
+ draft: main
3264
+ signal_name: concept_score
3265
+ - path: original-instruction
3266
+ signal:
3267
+ embedding: gte-small
3268
+ namespace: lilac
3269
+ concept_name: source-code
3270
+ version: null
3271
+ draft: main
3272
+ signal_name: concept_score
3273
+ - path: original-instruction
3274
+ signal:
3275
+ embedding: gte-small
3276
+ namespace: lilac
3277
+ concept_name: toxicity
3278
+ version: null
3279
+ draft: main
3280
+ signal_name: concept_score
3281
+ - path: original-context
3282
+ signal:
3283
+ embedding: gte-small
3284
+ min_cluster_size: 5
3285
+ umap_n_components: 10
3286
+ signal_name: cluster_hdbscan
3287
+ - path: original-response
3288
+ signal:
3289
+ embedding: gte-small
3290
+ min_cluster_size: 5
3291
+ umap_n_components: 10
3292
+ signal_name: cluster_hdbscan
3293
+ - path: original-response
3294
+ signal:
3295
+ embedding: gte-small
3296
+ namespace: lilac
3297
+ concept_name: legal-termination
3298
+ version: null
3299
+ draft: main
3300
+ signal_name: concept_score
3301
+ - path: original-response
3302
+ signal:
3303
+ embedding: gte-small
3304
+ namespace: lilac
3305
+ concept_name: negative-sentiment
3306
+ version: null
3307
+ draft: main
3308
+ signal_name: concept_score
3309
+ - path: original-response
3310
+ signal:
3311
+ embedding: gte-small
3312
+ namespace: lilac
3313
+ concept_name: non-english
3314
+ version: null
3315
+ draft: main
3316
+ signal_name: concept_score
3317
+ - path: original-response
3318
+ signal:
3319
+ embedding: gte-small
3320
+ namespace: lilac
3321
+ concept_name: positive-sentiment
3322
+ version: null
3323
+ draft: main
3324
+ signal_name: concept_score
3325
+ - path: original-response
3326
+ signal:
3327
+ embedding: gte-small
3328
+ namespace: lilac
3329
+ concept_name: profanity
3330
+ version: null
3331
+ draft: main
3332
+ signal_name: concept_score
3333
+ - path: original-response
3334
+ signal:
3335
+ embedding: gte-small
3336
+ namespace: lilac
3337
+ concept_name: question
3338
+ version: null
3339
+ draft: main
3340
+ signal_name: concept_score
3341
+ - path: original-response
3342
+ signal:
3343
+ embedding: gte-small
3344
+ namespace: lilac
3345
+ concept_name: source-code
3346
+ version: null
3347
+ draft: main
3348
+ signal_name: concept_score
3349
+ - path: original-response
3350
+ signal:
3351
+ embedding: gte-small
3352
+ namespace: lilac
3353
+ concept_name: toxicity
3354
+ version: null
3355
+ draft: main
3356
+ signal_name: concept_score
3357
+ - path:
3358
+ - new-instruction
3359
+ - value
3360
+ - '*'
3361
+ signal:
3362
+ embedding: gte-small
3363
+ min_cluster_size: 5
3364
+ umap_n_components: 10
3365
+ signal_name: cluster_hdbscan
3366
  - path:
3367
  - new-instruction
3368
  - value
 
3457
  - '*'
3458
  signal:
3459
  embedding: gte-small
3460
+ min_cluster_size: 5
3461
+ umap_n_components: 10
3462
+ signal_name: cluster_hdbscan
3463
  - path:
3464
  - new-response
3465
  - value
3466
  - '*'
3467
  signal:
3468
  embedding: gte-small
3469
+ min_cluster_size: 5
3470
+ umap_n_components: 10
3471
+ signal_name: cluster_hdbscan
3472
  - path:
3473
  - new-response
3474
  - value
 
4138
  signal_name: concept_score
4139
  - path: question
4140
  signal:
 
 
 
4141
  signal_name: cluster_dbscan
4142
  - path: response
4143
  signal:
 
 
 
4144
  signal_name: cluster_dbscan
4145
+ - path: question
4146
+ signal:
4147
+ embedding: gte-small
4148
+ min_cluster_size: 5
4149
+ umap_n_components: 10
4150
+ signal_name: cluster_hdbscan
4151
+ - path: response
4152
+ signal:
4153
+ embedding: gte-small
4154
+ min_cluster_size: 5
4155
+ umap_n_components: 10
4156
+ signal_name: cluster_hdbscan
4157
  settings:
4158
  ui:
4159
  media_paths:
 
4550
  - translation
4551
  - es
4552
  signal:
 
 
 
4553
  signal_name: cluster_dbscan
4554
  - path:
4555
  - translation
4556
  - en
4557
  signal:
 
 
 
4558
  signal_name: cluster_dbscan
4559
+ - path:
4560
+ - translation
4561
+ - es
4562
+ signal:
4563
+ embedding: gte-small
4564
+ min_cluster_size: 5
4565
+ umap_n_components: 10
4566
+ signal_name: cluster_hdbscan
4567
+ - path:
4568
+ - translation
4569
+ - en
4570
+ signal:
4571
+ embedding: gte-small
4572
+ min_cluster_size: 5
4573
+ umap_n_components: 10
4574
+ signal_name: cluster_hdbscan
4575
  settings:
4576
  ui:
4577
  media_paths:
 
4740
  signal_name: concept_score
4741
  - path: lecture
4742
  signal:
 
 
 
4743
  signal_name: cluster_dbscan
4744
+ - path: lecture
4745
+ signal:
4746
+ embedding: gte-small
4747
+ min_cluster_size: 5
4748
+ umap_n_components: 10
4749
+ signal_name: cluster_hdbscan
4750
  settings:
4751
  ui:
4752
  media_paths:
 
4912
  signal_name: concept_score
4913
  - path: text
4914
  signal:
 
 
 
4915
  signal_name: cluster_dbscan
4916
+ - path: text
4917
+ signal:
4918
+ embedding: gte-small
4919
+ min_cluster_size: 5
4920
+ umap_n_components: 10
4921
+ signal_name: cluster_hdbscan
4922
  settings:
4923
  ui:
4924
  media_paths:
 
5083
  signal_name: concept_score
5084
  - path: overview
5085
  signal:
 
 
 
5086
  signal_name: cluster_dbscan
5087
+ - path: overview
5088
+ signal:
5089
+ embedding: gte-small
5090
+ min_cluster_size: 5
5091
+ umap_n_components: 10
5092
+ signal_name: cluster_hdbscan
5093
  settings:
5094
  ui:
5095
  media_paths:
 
5405
  - outline
5406
  - '*'
5407
  signal:
 
 
 
5408
  signal_name: cluster_dbscan
5409
  - path:
5410
  - concepts
5411
  - '*'
5412
  signal:
 
 
 
5413
  signal_name: cluster_dbscan
5414
  - path: markdown
5415
  signal:
 
 
 
5416
  signal_name: cluster_dbscan
5417
+ - path:
5418
+ - outline
5419
+ - '*'
5420
+ signal:
5421
+ embedding: gte-small
5422
+ min_cluster_size: 5
5423
+ umap_n_components: 10
5424
+ signal_name: cluster_hdbscan
5425
+ - path:
5426
+ - concepts
5427
+ - '*'
5428
+ signal:
5429
+ embedding: gte-small
5430
+ min_cluster_size: 5
5431
+ umap_n_components: 10
5432
+ signal_name: cluster_hdbscan
5433
+ - path: markdown
5434
+ signal:
5435
+ embedding: gte-small
5436
+ min_cluster_size: 5
5437
+ umap_n_components: 10
5438
+ signal_name: cluster_hdbscan
5439
  settings:
5440
  ui:
5441
  media_paths:
docker_start.sh CHANGED
@@ -5,6 +5,6 @@ set -e
5
 
6
  lilac hf-docker-start
7
  gunicorn lilac.server:app \
8
- --bind 0.0.0.0:5432 \
9
  --preload -k uvicorn.workers.UvicornWorker \
10
  --timeout 120
 
5
 
6
  lilac hf-docker-start
7
  gunicorn lilac.server:app \
8
+ --bind 0.0.0.0:8000 \
9
  --preload -k uvicorn.workers.UvicornWorker \
10
  --timeout 120