{
 "source": "fastino/gliner2.5-base-v1@1a8bc24e00dc7300b9017c81d63e3dcdabb26596",
 "variant": "wfp16",
 "runtime": {
  "litert": "2.2.0 (validated on the Mac with ai-edge-litert 2.2.0; com.google.ai.edge.litert:litert:2.2.0 on Android)",
  "api": "CompiledModel",
  "converter": {
   "litert-torch": "0.9.4",
   "ai-edge-litert": "2.2.0",
   "ai-edge-quantizer": "0.9.0",
   "litert-converter": "0.4.0",
   "torch": "2.13.0",
   "transformers": "4.57.6",
   "gliner2": "2.0.0"
  }
 },
 "graphs": [
  {
   "path": "gliner2_base_encoder_wfp16.tflite",
   "role": "encoder",
   "variant": "wfp16",
   "shipped": true,
   "about": "encoder FULLY_CONNECTED weights stored float16 behind DEQUANTIZE (ai-edge-quantizer FLOAT_CASTING); FP32 activations, position tables and norms; finite -1e4 mask; head and scorer FP32",
   "bytes": 189419856,
   "signatures": [
    {
     "key": "t128",
     "inputs": [
      {
       "name": "attention_mask",
       "dtype": "float32",
       "shape": [
        1,
        128
       ]
      },
      {
       "name": "inputs_embeds",
       "dtype": "float32",
       "shape": [
        1,
        128,
        768
       ]
      }
     ],
     "outputs": [
      {
       "name": "hidden_states",
       "dtype": "float32",
       "shape": [
        1,
        128,
        768
       ]
      }
     ]
    }
   ],
   "use": "the CPU (XNNPACK weight cache), or the GPU at explicit FP32 precision"
  },
  {
   "path": "gliner2_base_encoder_fp16hw.tflite",
   "role": "encoder",
   "variant": "fp16hw",
   "shipped": false,
   "about": "wfp16's storage over the FP16-hardware encoder: LayerNorm power-of-two pre-scale (x*2^-k, eps*4^-k, k from the corpus' FP32 ranges), first feed-forward FULLY_CONNECTED split in two halves; head and scorer FP32",
   "bytes": 189416320,
   "signatures": [
    {
     "key": "t128",
     "inputs": [
      {
       "name": "attention_mask",
       "dtype": "float32",
       "shape": [
        1,
        128
       ]
      },
      {
       "name": "inputs_embeds",
       "dtype": "float32",
       "shape": [
        1,
        128,
        768
       ]
      }
     ],
     "outputs": [
      {
       "name": "hidden_states",
       "dtype": "float32",
       "shape": [
        1,
        128,
        768
       ]
      }
     ]
    }
   ],
   "use": "the GPU at FP16 precision (decisions as wfp16 at FP32)"
  },
  {
   "path": "gliner2_base_head_fp32.tflite",
   "role": "head",
   "about": "the boundary head after the encoder (FluidInference's Core ML features graph without its encoder): FP32, CPU",
   "bytes": 8517024,
   "signatures": [
    {
     "key": "t128",
     "inputs": [
      {
       "name": "cls_indices",
       "dtype": "int32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "cls_mask",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "hidden_states",
       "dtype": "float32",
       "shape": [
        1,
        128,
        768
       ]
      },
      {
       "name": "query_indices",
       "dtype": "int32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "query_mask",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "text_indices",
       "dtype": "int32",
       "shape": [
        1,
        64
       ]
      },
      {
       "name": "text_mask",
       "dtype": "float32",
       "shape": [
        1,
        64
       ]
      }
     ],
     "outputs": [
      {
       "name": "boundary_mask",
       "dtype": "float32",
       "shape": [
        1,
        65
       ]
      },
      {
       "name": "boundary_states",
       "dtype": "float32",
       "shape": [
        1,
        65,
        128
       ]
      },
      {
       "name": "classification_logits",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "count_log_rates",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "end_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "inside_prefix",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "inside_prefix_mean",
       "dtype": "float32",
       "shape": [
        1,
        8,
        1
       ]
      },
      {
       "name": "null_logits",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "pool_end_projection",
       "dtype": "float32",
       "shape": [
        1,
        65,
        128
       ]
      },
      {
       "name": "pool_start_projection",
       "dtype": "float32",
       "shape": [
        1,
        65,
        128
       ]
      },
      {
       "name": "query_states",
       "dtype": "float32",
       "shape": [
        1,
        8,
        768
       ]
      },
      {
       "name": "start_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "text_states",
       "dtype": "float32",
       "shape": [
        1,
        64,
        768
       ]
      }
     ]
    }
   ]
  },
  {
   "path": "gliner2_base_scorer_fp32.tflite",
   "role": "scorer",
   "about": "the candidate scorer and the record candidate states (FluidInference's Core ML scorer): FP32, CPU",
   "bytes": 2126928,
   "signatures": [
    {
     "key": "p192",
     "inputs": [
      {
       "name": "boundary_states",
       "dtype": "float32",
       "shape": [
        1,
        65,
        128
       ]
      },
      {
       "name": "candidate_compatibility",
       "dtype": "float32",
       "shape": [
        1,
        192
       ]
      },
      {
       "name": "candidate_indices",
       "dtype": "int32",
       "shape": [
        1,
        192,
        2
       ]
      },
      {
       "name": "candidate_mask",
       "dtype": "float32",
       "shape": [
        1,
        192
       ]
      },
      {
       "name": "end_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "inside_prefix",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "inside_prefix_mean",
       "dtype": "float32",
       "shape": [
        1,
        8,
        1
       ]
      },
      {
       "name": "query_mask",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "query_states",
       "dtype": "float32",
       "shape": [
        1,
        8,
        768
       ]
      },
      {
       "name": "start_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "text_mask",
       "dtype": "float32",
       "shape": [
        1,
        64
       ]
      },
      {
       "name": "text_states",
       "dtype": "float32",
       "shape": [
        1,
        64,
        768
       ]
      }
     ],
     "outputs": [
      {
       "name": "candidate_states",
       "dtype": "float32",
       "shape": [
        1,
        192,
        768
       ]
      },
      {
       "name": "pair_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        192
       ]
      }
     ]
    }
   ]
  },
  {
   "path": "gliner2_base_explicit_fp32.tflite",
   "role": "explicit",
   "about": "explicit-span scorer: the boundary proposer's prior and the pair scorer over up to 64 forced spans per query (attributes, choice fields); FP32, CPU",
   "bytes": 2051844,
   "signatures": [
    {
     "key": "w64_q8_s64",
     "inputs": [
      {
       "name": "text_states",
       "dtype": "float32",
       "shape": [
        1,
        64,
        768
       ]
      },
      {
       "name": "text_mask",
       "dtype": "float32",
       "shape": [
        1,
        64
       ]
      },
      {
       "name": "query_states",
       "dtype": "float32",
       "shape": [
        1,
        8,
        768
       ]
      },
      {
       "name": "query_mask",
       "dtype": "float32",
       "shape": [
        1,
        8
       ]
      },
      {
       "name": "boundary_states",
       "dtype": "float32",
       "shape": [
        1,
        65,
        128
       ]
      },
      {
       "name": "start_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "end_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "inside_prefix",
       "dtype": "float32",
       "shape": [
        1,
        8,
        65
       ]
      },
      {
       "name": "inside_prefix_mean",
       "dtype": "float32",
       "shape": [
        1,
        8,
        1
       ]
      },
      {
       "name": "span_indices",
       "dtype": "int32",
       "shape": [
        1,
        8,
        64,
        2
       ]
      },
      {
       "name": "span_mask",
       "dtype": "float32",
       "shape": [
        1,
        8,
        64
       ]
      }
     ],
     "outputs": [
      {
       "name": "span_logits",
       "dtype": "float32",
       "shape": [
        1,
        8,
        64
       ]
      }
     ]
    }
   ]
  },
  {
   "path": "gliner2_base_relation_fp32.tflite",
   "role": "relation",
   "about": "relation scorer over up to 256 head/tail pairs of up to 4 relation types (pairs chosen by gliner2's typed pair generator); FP32, CPU",
   "bytes": 23658524,
   "signatures": [
    {
     "key": "w64_r4_p256",
     "inputs": [
      {
       "name": "text_states",
       "dtype": "float32",
       "shape": [
        1,
        64,
        768
       ]
      },
      {
       "name": "text_length",
       "dtype": "int32",
       "shape": [
        1
       ]
      },
      {
       "name": "relation_states",
       "dtype": "float32",
       "shape": [
        1,
        4,
        1536
       ]
      },
      {
       "name": "batch_index",
       "dtype": "int32",
       "shape": [
        256
       ]
      },
      {
       "name": "relation_index",
       "dtype": "int32",
       "shape": [
        256
       ]
      },
      {
       "name": "head_start",
       "dtype": "int32",
       "shape": [
        256
       ]
      },
      {
       "name": "head_end",
       "dtype": "int32",
       "shape": [
        256
       ]
      },
      {
       "name": "tail_start",
       "dtype": "int32",
       "shape": [
        256
       ]
      },
      {
       "name": "tail_end",
       "dtype": "int32",
       "shape": [
        256
       ]
      },
      {
       "name": "pair_mask",
       "dtype": "float32",
       "shape": [
        256
       ]
      }
     ],
     "outputs": [
      {
       "name": "relation_logits",
       "dtype": "float32",
       "shape": [
        256
       ]
      }
     ]
    }
   ]
  },
  {
   "path": "gliner2_base_record_assignment_fp32.tflite",
   "role": "record_assignment",
   "about": "record decoder: instance-to-field assignment over 8 fields with up to 192 candidates each, for up to 1536 instances, plus object and latent-seed scores; FP32, CPU",
   "bytes": 1193288,
   "signatures": [
    {
     "key": "f8_c192_i1536",
     "inputs": [
      {
       "name": "instance_states",
       "dtype": "float32",
       "shape": [
        1536,
        768
       ]
      },
      {
       "name": "field_queries",
       "dtype": "float32",
       "shape": [
        8,
        768
       ]
      },
      {
       "name": "field_candidate_states",
       "dtype": "float32",
       "shape": [
        8,
        192,
        768
       ]
      }
     ],
     "outputs": [
      {
       "name": "assignment_logits",
       "dtype": "float32",
       "shape": [
        1536,
        8,
        193
       ]
      },
      {
       "name": "object_logits",
       "dtype": "float32",
       "shape": [
        1536
       ]
      },
      {
       "name": "latent_seed_logits",
       "dtype": "float32",
       "shape": [
        1536
       ]
      }
     ]
    }
   ]
  },
  {
   "path": "gliner2_base_record_anchorless_fp32.tflite",
   "role": "record_anchorless",
   "about": "record decoder: its 32 learned instances attending over up to 1536 field candidates (anchorless records); FP32, CPU",
   "bytes": 2875604,
   "signatures": [
    {
     "key": "c1536",
     "inputs": [
      {
       "name": "context_states",
       "dtype": "float32",
       "shape": [
        1536,
        768
       ]
      },
      {
       "name": "context_mask",
       "dtype": "float32",
       "shape": [
        1536
       ]
      }
     ],
     "outputs": [
      {
       "name": "instance_states",
       "dtype": "float32",
       "shape": [
        32,
        768
       ]
      }
     ]
    }
   ]
  }
 ],
 "table": {
  "path": "gliner2_base_embeddings_f16.bin",
  "dtype": "float16",
  "byteOrder": "little-endian",
  "shape": [
   128011,
   768
  ],
  "layout": "row-major, no header: row `id` is bytes [id * 1536, id * 1536 + 1536), 768 float16 values",
  "bytes": 196624896,
  "sha256": "22c2e937cfc17e1ee9c8f3ec6f1449d44dc9230cfcbb8737c9bae8492628a3e8",
  "use": "memory-map the file; for each of the 128 positions take the row of its zero-padded id (padding takes id 0's row) and upcast to float32 into inputs_embeds"
 },
 "bucket": {
  "tokens": 128,
  "words": 64,
  "queries": 8,
  "choices": 8,
  "pool": 192,
  "boundaries": 65,
  "about": "the host pads to T=128 tokens, W=64 words, Q=8 queries, C=8 choices, P=192 candidates and refuses larger requests"
 },
 "host": {
  "steps": "as the gliner2 library does: pack and pad the request, then encoder -> head -> the shared candidate pool on the host -> scorer -> thresholds, flat overlap resolution and spans; entities abstain when sigmoid(null_logits) > 0.5",
  "changes": [
   "input_ids no longer enter a graph: the host looks them up in the float16 table into inputs_embeds",
   "attention_mask is float32 (1.0 / 0.0) instead of int32",
   "FluidInference's features graph is two graphs: encoder (inputs_embeds, attention_mask -> hidden_states) and head (hidden_states plus the six other feature inputs -> the same thirteen outputs)",
   "the scorer, explicit-span, relation and record graphs take FluidInference's inputs and return its outputs, FP32",
   "bind inputs by name; signatures: encoder `t128`, head `t128`, scorer `p192`, explicit `w64_q8_s64`, relation `w64_r4_p256`, record assignment `f8_c192_i1536`, anchorless records `c1536`"
  ]
 },
 "accelerators": {
  "cpu": "encoder on XNNPACK with the weight cache path set (2 threads in the validation); head and scorer on the CPU (1 thread each)",
  "gpu": "encoder only (the head and scorer stay on the CPU): explicit precision (FP32 with wfp16, FP16 with fp16hw) and constant tensor sharing"
 },
 "tokenizer": {
  "file": "tokenizer/tokenizer.json",
  "specialTokens": {
   "[PAD]": 0,
   "[CLS]": 1,
   "[SEP]": 2,
   "[UNK]": 3,
   "[MASK]": 128000,
   "[SEP_STRUCT]": 128001,
   "[SEP_TEXT]": 128002,
   "[P]": 128003,
   "[C]": 128004,
   "[E]": 128005,
   "[R]": 128006,
   "[L]": 128007,
   "[EXAMPLE]": 128008,
   "[OUTPUT]": 128009,
   "[DESCRIPTION]": 128010
  }
 },
 "provenance": {
  "graphs": "encoder graph code ported from litert-community/GLiNER2.5-Decide-LiteRT (Apache-2.0); head and scorer wrappers adapted from FluidInference/gliner2-5-base-coreml (Apache-2.0)"
 },
 "referenceRuntime": "runtime/litert_boundary_runtime.py (gliner2 2.0.0's BoundaryExtractor over these graphs)"
}