Create common_codec.sh‎

Author: Feilong607

eval_encoder/shells_eval_ap/common_codec.sh

@@ -0,0 +1,157 @@
+#!/bin/bash
+# ============================================================================
+# Common Configuration Script - for attentive_probe_codec evaluation
+# Usage: Source this file in model scripts, then call run_attentive_probe_codec
+# ============================================================================
+
+# Environment setup
+export PYTHONPATH=../
+
+# Default dataset list (can be overridden in calling script)
+DEFAULT_DATASETS=(
+    "ssv2"
+    "diving48"
+    "perception_test"
+    "epic_verb"
+    "epic_noun"
+    "hmdb51"
+    "k400"
+    "charadesego"
+)
+
+# ============================================================================
+# Get batch size based on dataset name
+# Args: $1 - dataset name
+# ============================================================================
+get_batch_size() {
+    local dataset="$1"
+    if [[ "$dataset" == "ssv2" || "$dataset" == "diving48" || "$dataset" == "perception_test" ]]; then
+        echo 4
+    elif [[ "$dataset" == "hmdb51" ]]; then
+        echo 2
+    else
+        echo 16
+    fi
+}
+
+# ============================================================================
+# Get epochs based on dataset name
+# Args: $1 - dataset name
+# ============================================================================
+get_epochs() {
+    local dataset="$1"
+    if [[ "$dataset" == "hmdb51" ]]; then
+        echo 30
+    elif [[ "$dataset" == "diving48" ]]; then
+        echo 30
+    elif [[ "$dataset" == "perception_test" ]]; then
+        echo 30
+    else
+        echo 10
+    fi
+}
+
+# ============================================================================
+# Get codec parameters based on dataset name
+# Args: $1 - dataset name
+# Returns: Sets CODEC_MV_COMPENSATE, CODEC_STATIC_ABS_THRESH, CODEC_STATIC_REL_THRESH
+# ============================================================================
+get_codec_params() {
+    local dataset="$1"
+    if [[ "$dataset" == "diving48" || "$dataset" == "perception_test" ]]; then
+        # Parameters for diving48 and perception_test
+        CODEC_MV_COMPENSATE="similarity"
+        CODEC_STATIC_ABS_THRESH="126"
+        CODEC_STATIC_REL_THRESH="0.38"
+    else
+        # Parameters for other datasets
+        CODEC_MV_COMPENSATE="similarity"
+        CODEC_STATIC_ABS_THRESH="116"
+        CODEC_STATIC_REL_THRESH="0.55"
+    fi
+}
+
+# ============================================================================
+# Run attentive_probe_codec evaluation
+# Required variables to set before calling:
+#   - MODEL_FAMILY: model family (required)
+#   - MODEL_NAME: model name (required)
+#   - MODEL_WEIGHT: model weight path (optional, default "NULL")
+#   - FRAMES_TOKEN_NUM: token count (optional, default 196)
+#   - EMBEDDING_SIZE: embedding dimension (optional, default 768)
+#   - INPUT_SIZE: input size (optional, not passed if unset)
+#   - NUM_FRAMES: number of frames (optional, not passed if unset)
+#   - DATASETS: dataset array (optional, uses DEFAULT_DATASETS if unset/empty)
+#   - REPORT_DIR_SUFFIX: report directory suffix (optional, e.g. "_64frames_codec")
+# ============================================================================
+run_attentive_probe_codec() {
+    # Set default values
+    MODEL_WEIGHT="${MODEL_WEIGHT:-NULL}"
+    FRAMES_TOKEN_NUM="${FRAMES_TOKEN_NUM:-196}"
+    EMBEDDING_SIZE="${EMBEDDING_SIZE:-768}"
+    REPORT_DIR_SUFFIX="${REPORT_DIR_SUFFIX:-}"
+
+    # Use custom datasets or default datasets
+    if [[ -z "${DATASETS+x}" ]] || [[ ${#DATASETS[@]} -eq 0 ]]; then
+        DATASETS=("${DEFAULT_DATASETS[@]}")
+    fi
+
+    # Build report directory
+    BASE_REPORT_DIR="result_attentive_probe/${MODEL_FAMILY}/${MODEL_NAME}${REPORT_DIR_SUFFIX}"
+
+    # Loop through each dataset for testing
+    for DATASET in "${DATASETS[@]}"; do
+        BATCH_SIZE=$(get_batch_size "$DATASET")
+        EPOCHS=$(get_epochs "$DATASET")
+        
+        # Get codec-specific parameters for this dataset
+        get_codec_params "$DATASET"
+
+        echo "DATASET=$DATASET, BATCH_SIZE=$BATCH_SIZE"
+        echo "Codec params: mv_compensate=${CODEC_MV_COMPENSATE}, static_abs_thresh=${CODEC_STATIC_ABS_THRESH}, static_rel_thresh=${CODEC_STATIC_REL_THRESH}"
+
+        echo "========================================================"
+        echo "Start testing dataset: ${DATASET}"
+        echo "Model: ${MODEL_NAME}"
+        echo "Batch Size: ${BATCH_SIZE}"
+        echo "Report Dir: ${BASE_REPORT_DIR}/${DATASET}"
+        echo "========================================================"
+
+        # Build output directory
+        SAVE_DIR="${BASE_REPORT_DIR}/${DATASET}"
+        mkdir -p "$SAVE_DIR"
+
+        # Build extra arguments
+        EXTRA_ARGS=""
+        if [[ -n "${INPUT_SIZE}" ]]; then
+            EXTRA_ARGS="${EXTRA_ARGS} --input_size ${INPUT_SIZE}"
+        fi
+        if [[ -n "${NUM_FRAMES}" ]]; then
+            EXTRA_ARGS="${EXTRA_ARGS} --num_frames ${NUM_FRAMES}"
+        fi
+
+        torchrun --nproc_per_node 8 --master_port 15555 \
+            attentive_probe_codec.py \
+            --eval_freq 1 \
+            --default_lr_list 0.0001 \
+            --default_epoch "${EPOCHS}" \
+            --batch_size ${BATCH_SIZE} \
+            --default_weight_decay 0 \
+            --dali_py_num_workers 8 \
+            --model_family "${MODEL_FAMILY}" \
+            --model_name "${MODEL_NAME}" \
+            --model_weight "${MODEL_WEIGHT}" \
+            --dataset "${DATASET}" \
+            --save_report "${SAVE_DIR}" \
+            --frames_token_num ${FRAMES_TOKEN_NUM} \
+            --embedding_size ${EMBEDDING_SIZE} \
+            --mv_compensate ${CODEC_MV_COMPENSATE} \
+            --static_abs_thresh ${CODEC_STATIC_ABS_THRESH} \
+            --static_rel_thresh ${CODEC_STATIC_REL_THRESH} \
+            --static_fallback 1 \
+            ${EXTRA_ARGS}
+
+        echo "Finished testing ${DATASET}"
+        echo ""
+    done
+}